Qualitative Methods in Social Work Research

Third Edition

2
3
Qualitative Methods in Social Work Research
Third Edition

Deborah K. Padgett
New York University

4
FOR INFORMATION:

SAGE Publications, Inc.

2455 Teller Road

Thousand Oaks, California 91320

E-mail: order@sagepub.com

SAGE Publications Ltd.

1 Oliver’s Yard

55 City Road

London, EC1Y 1SP

United Kingdom

SAGE Publications India Pvt. Ltd.

B 1/I 1 Mohan Cooperative Industrial Area

Mathura Road, New Delhi 110 044

India

SAGE Publications Asia-Pacific Pte. Ltd.

3 Church Street

#10-04 Samsung Hub

Singapore 049483

Copyright © 2017 by SAGE Publications, Inc.

All rights reserved. No part of this book may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying, recording, or by any information storage and retrieval
system, without permission in writing from the publisher.
Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Names: Padgett, Deborah, author.

Title: Qualitative methods in social work research / Deborah K. Padgett, New York University, New York.

Description: Third edition. | Los Angeles : SAGE, [2017] | Includes bibliographical references and index.

Identifiers: LCCN 2015045298 | ISBN 9781452256702 (pbk. : alk. paper)

Subjects: LCSH: Social service—Research—Methodology.

Classification: LCC HV11 .P24 2017 | DDC 361.3072/1—dc23 LC record available at http://lccn.loc.gov/2015045298

This book is printed on acid-free paper.

Acquisitions Editor: Nathan Davidson

Editorial Assistant: Carrie Montoya

5
Production Editor: Bennie Clark Allen

Copy Editor: Michelle Ponce

Typesetter: C&M Digitals (P) Ltd.

Proofreader: Rae-Ann Goodwin

Indexer: Jean Casalegno

Cover Designer: Janet Kiesel

Marketing Manager: Shari Countryman

6
Brief Contents
1. List of Tables, Figures, and Boxes
2. Preface to the Third Edition
3. Acknowledgments
4. 1. Qualitative Methods in Context
5. 2. Choosing the Right Qualitative Approach(es)
6. 3. Designing the Study and Getting Started
7. 4. Ethical and Emotional Issues in Qualitative Research
8. 5. Data Collection: Observation, Interviewing, and Use of Documents
9. 6. Data Analysis
10. 7. Interpretation: Recognizing Salience
11. 8. Strategies for Rigor and Trustworthiness
12. 9. Telling the Story: Writing the Qualitative Study
13. 10. Mixed Methods
14. Appendix: Writing a Qualitative Methods Proposal for External Funding
15. References
16. Index
17. About the Author

7
Detailed Contents
List of Tables, Figures, and Boxes
Preface to the Third Edition
Acknowledgments
1. Qualitative Methods in Context
Differences From and Similarities With Quantitative Methods
The Intellectual Life Course of Qualitative Methods
The Formative Years
The Rise of Quantification and Operationalism
Epistemological Differences Among Qualitative Researchers
The Pragmatic Middle Ground
Evidence-Based Practice and Implementation Science
Theoretical and Conceptual Frameworks in Qualitative Inquiry
Some Are a Better Fit Than Others
Analytic Induction and Abduction: Nondeductive Thinking
The Importance of Concepts
The Place and Timing of Theories in Qualitative Studies
Reasons for Doing Qualitative Research
Desirable Qualities and Skills in the Qualitative Researcher-as-Instrument
Studying the Familiar Versus the Unfamiliar
Values and Social Responsibility in Qualitative Research
Introducing the New York Services Study (NYSS) and the New York Recovery Study (NYRS)
Summary and Concluding Thoughts
Exercises
Additional Readings
Journals That Feature Qualitative Studies and Methods
Websites for Qualitative Methods
Selected Monographs
2. Choosing the Right Qualitative Approach(es)
Six Primary Approaches in Qualitative Research
Ethnography
Grounded Theory
Case Study Analysis
Narrative Approaches
Phenomenological Analysis
Action and Community-Engaged Research
The Six Approaches Revisited: Change Over Time
Mixing and Matching Qualitative Approaches: Risks and Benefits

8
 Introducing Strategies for Rigor and Trustworthiness
Qualitative Methods in Program Evaluation and Implementation Research
Summary and Concluding Thoughts
Exercises
Additional Readings
Websites Related to CBPR
3. Designing the Study and Getting Started
Prologue: Addressing Ethical Considerations Early On
Flexible and Iterative Designs
Choosing the Topic and Making the Argument
Reviewing the Literature
Developing a Conceptual Framework
Formulating Research Questions
Designing the Study
Questions Posed (and Needing Answers) by Qualitative Research Designs
The Element of Time
Specific Aspects of Qualitative Designs
Using Multiple Qualitative Methods at Different Levels
Sampling and Selection Strategies
Sample Size Considerations
Recruiting and Retaining Study Participants
How Many Interviews? Issues of Quantity and Quality
Getting Permission(s) and Obtaining Rapport
Presentation of Self
Research Self-Disclosure: How Much is Enough? Too Much?
Issues of Identity: Gender, Age, Race/Ethnicity, and Social Class
Summary and Concluding Thoughts
Exercises
Additional Readings
4. Ethical and Emotional Issues in Qualitative Research
Deception and Disclosure
Informed Consent
Coercion and “Deformed” Consent
Confidentiality and Privacy
Distress and Emotional Harm
Incentives, Payback, and Maintaining Goodwill
Institutional Review Boards and Qualitative Research
Dealing With Moral Ambiguity and Risk
Ethical Issues in Community-Engaged Research
Ethical Issues in Cross-National Research

9
 Taking It Personally: Emotions in Qualitative Research
Socially Responsible Research as Ethical Research
Summary and Concluding Thoughts
Exercises
Additional Readings
5. Data Collection: Observation, Interviewing, and Use of Documents
Observation in Qualitative Studies
“Doing” Observation: The Best (and Only True) Way to Learn
Recording Observational Data
Uses of Video and Photography
In-Depth Interviewing
Critiques of Interviewing: Realism and Authenticity
Informal Field Interviews
Focus Group Interviews
Individual Interviews
Linked Interviews
Interviewing Children and Other Vulnerable Populations
Key Informant, Elite, and Expert Interviews
Developing the Interview Guide
Elicitation Techniques in Qualitative Interviews
Interview Guides for Second (and Third) Interviews
A Few Guidelines for Starting Out
Conducting the Interview: The Importance of Probes
After the Interview
Matching Interviewers to Respondents: The Effects of Age, Gender, Race, and Other
Characteristics
Common Problems and Errors in Qualitative Interviewing
Emotional Issues in Interviewing
Interviewing in a Non-English Language
Interviewing Across the Digital Landscape
Use of Audio and Video Recorders
Using Documents and Other Archival Materials
Ending Data Collection
Summary and Concluding Thoughts
Exercises
Additional Readings
6. Data Analysis
Data Management: Dealing With Volume Early On
Using Qualitative Data Analysis (QDA) Software
Teams, Multiple Users, and Cross-Site Databases

10
 Transcribing Interviews
Translating and Transcribing in a Non-English Language
Qualitative Data Analysis: Beginning The Search for Meaning
Case Summaries
Theories and Concepts in Qualitative Data Analysis
Data Analysis in Diverse Qualitative Approaches
Content Analysis
Ethnography and Data Analysis
Case Study Analysis
Data Analysis in Narrative Approaches
Phenomenological Analyses
Action and Community-Based Participatory Research
Analysis in Longitudinal Designs
Analyzing Visual Data
Coding
Varied Approaches to Coding and Co-Coding
Starting Out: Identifying Code-Worthy Material
Labeling Coded Material and Developing a Codebook
Documenting Coding Procedures
Comparing and Contrasting: Memo Writing as Forward Motion
Secondary Analyses of Qualitative Data
Hybrid and Mixed Qualitative Approaches
Summary and Concluding Thoughts
Exercises
Additional Readings
QDA Software Resources
7. Interpretation: Recognizing Salience
The Role of the Researcher(s)
Framing Devices to Assist in Interpretation
Using Qualitative Data Analysis (QDA) Software for Interpretation
Saturation
Bringing Theor(ies) in
An Example of a Conceptual Schema
Salience Through Absence: What Is Not Said (or Seen)
Feminist and Other Critiques in Interpretation
Rhetorical Devices
Negative Case Analysis and Disconfirmation
The Six Approaches Revisited
Qualitative Meta-Synthesis as Meta-Interpretation
In Pursuit of Methodological Transparency

11
 Summary and Concluding Thoughts
Exercises
Additional Readings
8. Strategies for Rigor and Trustworthiness
Evaluating Qualitative Research
Evaluative Criteria and External Standards
In the Eye of the Beholder?
Attending to Rigor and Trustworthiness
The Meaning(s) of Generalizability in Qualitative Methods
Threats to the Trustworthiness of Qualitative Studies
Strategies for Rigor in Qualitative Research
Prolonged Engagement
Triangulation of Data
Peer Debriefing and Support
Member Checking
Negative Case Analysis
Auditing—Leaving a Decision Trail
Strategies for Rigor Applied to the Six Qualitative Approaches
Dilemmas of Rigor in Action-Oriented Research
Appraising Quality: The Rise of Checklists,Criteria, and Guidelines
In Search of the Big Tent: Balancing Rigor and Relevance
Summary and Concluding Thoughts
Exercises
Additional Readings
9. Telling the Story: Writing the Qualitative Study
Preparing to Write: Key Decision Points
Identifying the Audience
The Researcher’s Role in the Report
Using Numbers in the Report
Authorship and Coauthorship Decisions
Authorship in and From Dissertations: A Negotiation With Few Rules
Organizing the Report—Key Components
Qualitative Reports Across the Diverse Approaches
Balancing Description and Interpretation
Writing Stance and Style
Rhetorical Devices Suited to Qualitative Studies
Writing Style: Rhythm and Grab
Use of Metaphors and Other Tropes
Challenges and Rewards of Publishing
Criteria for Review: The Qualitative Dilemma

12
 Summary and Concluding Thoughts
Exercises
Additional Readings
10. Mixed Methods
The Epistemology Question in Mixed Methods
The Rise of Mixed Methods and Their Rationale(s)
Integrating Mixed Methods
Types of Mixed Methods Designs
Sequential Designs
Concurrent Designs
Complicated Designs
Mixed Methods: Ways of Going About It
Structural and Design Decisions: What, When, and How?
Incorporating Existing Measures Into Qualitative Interviews
Quantifying Qualitative Data
Examples of Mixed Methods Studies: Implementation and Action Research
Writing the Mixed Methods Report
Challenges for Mixed Methods Studies
Mixed Method Approaches to Program and Practice Evaluation
Summary and Concluding Thoughts
Exercises
Additional Readings
Appendix: Writing a Qualitative Methods Proposal for External Funding
Finding the Right Fit With Funding Priorities
Getting Past the Catch–22
Setting the Stage
Questions From Proposal Reviewers
Budgeting Time and Resources
Twenty Tips for Writing the Proposal
Additional Readings
References
Index
About the Author

13
14
List of Tables, Figures, and Boxes

Tables
6.1 Initial Coding and Memos: An Example From the NYRS 167
6.2 Example of Codebook From the NYRS 171

15
Figures
1.1 The Foundations and Processes of a Qualitative Study 15
6.1 Screenshot From ATLAS.ti: Excerpt From Transcript in the NYRS 144
6.2 Screenshot From ATLAS.ti: Excerpt From Photo-Elicitation Interview in the NYRS 162
7.1 Themes and Subthemes From Phase 1 of the NYSS 191
8.1 Strategies for Enhancing Rigor and Trustworthiness 217
10.1 Mixed Methods Using Quantified Qualitative Data in the NYRS 269
10.2 An Oscillating Mixed Methods Design 272

16
Boxes
1.1 Social Work Research and Qualitative Methods 4
2.1 Compassionate Confinement: Ethnography in a Juvenile Detention Center 32
2.2 Surviving Childhood Sexual Abuse: An Exemplary Grounded Theory Study 35
2.3 Introducing an Evidence-Based Program Into a Hostile Setting: A Case Study Using Ethnographic
Methods 38
2.4 The Price of Not Knowing the Community: A Failed Trial of an HIV Prophylactic Product 44
3.1 Questions Posed by Qualitative Research Designs 63
3.2 Lost in Translation? Design Considerations for Non-English Language Studies 66
4.1 Moral Quandaries in an Ethnographic Study 89
5.1 Field Notes From a Client Meeting at a Transitional Housing Program in the NYRS 103
5.2 Photo-Elicitation in the NYRS 106
5.3 Preparing for a Focus Group Interview: The Need to Multitask 113
5.4 Tips for Qualitative Interviewing 114
5.5 Examples of Domains From the NYRS Interview Guide 118
5.6 The Importance of Probes Both Planned and Spontaneous: “Air Theory” and Breast Cancer 123
5.7 An Example of Interviewer Feedback From the NYRS 125
5.8 Internet Research on Persons With Early-Onset Alzheimer’s Disease 132
6.1 A Brief Quiz on Transcription “Errors” 147
6.2 Notation in Conversation Analysis: A Detailed Breakdown of Text
6.3 Example of a Memo for Interview Excerpt in Table 6.1 168
7.1 Operationalizing Saturation: A Study of Sex Workers in West Africa 187
7.2 The Life of a Hard-Working Code: Linking “Up and Out” to Theory 195
7.3 An Example of Quantified Thematic Development 201
7.4 Family Experiences of Child Death in a Pediatric Intensive Care Unit (PICU): A Meta-Synthesis
204
10.1 A Mixed Method Study (QUAN+qual) Testing a Measure of Social Capital in Peru and Vietnam
260
10.2 Strange Bedfellows? Embedding Qualitative Methods Within a Randomized Controlled Trial
(RCT) 262
10.3 “Quantitizing” Data in a Grounded Theory Study of Breast and Prostate Cancer Online
Discussion Boards 265
10.4 Quantifying Qualitative Data in the NYRS: Measuring Trajectories of Recovery 268
10.5 Serendipitous Findings and Mixed Methods: An Example From the Harlem Mammogram Study
274

17
18
Preface to the Third Edition

With each new edition of this book, the task becomes more complicated, yet it is worth every minute (and
hour). With an unprecedented rise in the level of sophistication and diversification, qualitative methods have
attracted attention far and wide. The internal affairs of qualitative inquiry are still boisterous, with paradigm
differences and a lack of consensus on definitions, terminologies, and procedures. The first edition of this
book opted for a generic approach, and the second edition took a broader and more nuanced view of the
diversification of the methods. As can be seen in the section below, this edition has gone several steps further
in revising and updating content.

19
What Is New in This Edition
An updated discussion of the paradigm debates
More attention to each of the six approaches throughout (ethnography, grounded theory, case study
analysis, narrative approaches, phenomenological approaches, and action-oriented research)
Early introductions to ethical issues and strategies for rigor (but also retaining separate chapters for
each)
More information on community-engaged, participatory action, and community-based participatory
research
All data collection methods consolidated into one chapter (Chapter 5)
More discussion of visual and ethnographic methods including photo-elicitation interviewing
Coding examples and further instruction in coding and thematic development
More on mixing qualitative approaches within a study while retaining and updating the chapter on
mixed (quantitative-qualitative) methods
More on QDA software including screenshots from ATLAS.ti
More discussion of ethical issues in cross-cultural and cross-national research
More discussion of the role of non-English languages in qualitative research
More advice on publishing qualitative studies (while retaining the Appendix on writing a qualitative
methods grant proposal)
Introductions to meta-syntheses and implementation science and their implications for qualitative
methods
A separate chapter dedicated to the interpretation of qualitative data (Chapter 7)
Discussion of current debates over criteria and checklists to assess the quality of qualitative research
Examples from the New York Recovery Study led by the author

20
Continuities From the Earlier Editions
Emphasis on six of the most commonly used qualitative methods (ethnography, grounded theory, case
studies, narrative approaches, phenomenological analysis, and action-oriented research)
Inclusion of specific illustrative examples from the literature as well as from my own research
A broad, multidisciplinary perspective that is rooted in social work but applicable to other practice
professions
An appreciation for the global nature of research problems and methods
Detailed guidelines on writing a qualitative research proposal in the Appendix
The importance of strategies for rigor in qualitative methods

21
Background to This Edition
The past 2 decades have been a time of gratifying professional growth for me as a qualitative researcher.
Having a doctorate in anthropology, subsequent postdoctoral training in public health, and federally funded
mixed method research on breast cancer and mammography set the stage for this book’s first edition. The
second edition came on the heels of a new development. In 2004, I was able to secure funding for a 4-year,
all-qualitative R01 grant from the National Institute of Mental Health (NIMH), a project titled the “New
York Services Study” (NYSS) that was mined for examples in the second edition. In 2010, lightning struck
again in a fortuitous way with the funding of another qualitative yet more ambitious methodological
undertaking made possible again by the support of NIMH. The New York Recovery Study, with its emphasis
on mental health recovery and adversity among formerly homeless adults with serious mental illness, repeated
some of the design features of the NYSS but with added ethnographic and visual methods including
shadowing (“go-along”) interviews, ethnographic site visits, and photo elicitation. These methods have
opened the door to thinking in new ways both spatially and visually. This third edition has provided an
opportunity to share our experiences with these methods and the benefits they can bring.

The reach of the “qualitative revolution” is both expansive and inclusive. Qualitative methods have become
increasingly visible in scholarly scientific journals and conference programs of the professions—education,
nursing, social work, public health, occupational therapy, and so on. They have become a multidisciplinary
meeting ground for researchers who want their methods to engage as well as elicit, to go deep rather than
broad and to incorporate social values sooner rather than later. Decidedly low tech for much of their existence,
qualitative studies have ridden the wave of technology with the development of varied analytic software
programs, digital audio and video recording, and the creative use of what the Internet has to offer in data
collection and file sharing. The creativity and commitment of qualitative researchers ensure that the methods
will stay current with the pace of technological change. Yet there is also the reassuring continuity of being able
to conduct a qualitative study with little more than an audio recorder and the passion to see it through.

Events in the world have brought home the need for responsive and agile research methods. A broad range of
problems from deeply embedded structural inequalities to multiplying world crises demand pragmatic
responses by researchers who can be flexible and responsive to local conditions. In addition, the multilevel
nature of “wicked problems” demands sophistication and methodological pluralism (Mertens, 2015). Broad-
based aggregate data and quantitative analyses are needed to address these problems, but grass-roots
perspectives from (and involvement with) those most deeply affected are essential to understanding and acting
to resolve these problems. Qualitative methods provide essential techniques for doing this.

22
Scope and Organization of the Book
This edition continues to emphasize the how-to aspects of qualitative inquiry. As the reader will see in
Chapter 1 and beyond, the terms qualitative methods and qualitative inquiry cover a variety of approaches as
well as philosophical stances. The chapters are roughly organized around the sequential steps involved in
carrying out a qualitative study—although the reader is cautioned that qualitative inquiry is recursive, going
back and forth between data and analyses and between analyses and write-up until findings become
consolidated into a form of representation that “fits” the phenomenon being studied.

Chapter 1 gives a brief history of the theoretical and disciplinary origins of qualitative methods, grounds the
reader in epistemological and other sources of diversity within the qualitative methods family, and sets the
stage for what follows. Chapter 2 includes detailed descriptions of six of the most commonly used qualitative
approaches (ethnography, case studies, grounded theory, narrative, phenomenological, and action-oriented
methods) to give the reader an overview of options from which to choose. Chapters 3 through 7 take the
reader through the various stages of carrying out qualitative research: study design and sampling (Chapter 3),
ethical issues (Chapter 4), data collection (Chapter 5), data analysis (Chapter 6), and interpretation (Chapter
7). Chapter 8 is devoted to rigor in qualitative research with attention to the contested nature of standards for
quality. Chapter 9 provides guidelines for writing up the study and telling the story. Chapter 10 is devoted to
mixed method studies, whose various iterations represent one of the fastest growing types of research. Finally,
the Appendix presents the basics of writing a qualitative methods proposal based on my experiences and the
experiences of others.

23
A Few Words on Writing and Terminology
This Preface seems the best place to make a few points about writing style and use of terminology. First, I
have tried to incorporate sensitivity to gender, ethnic, and other forms of diversity throughout the text.
Second, the awkwardness of nonsexist usage has led me to alternate use of masculine and feminine pronouns
and to use plural pronouns on other occasions. Finally, I have a preference for the term study participant with
occasional substitutions of respondent, interviewee, or informant, depending on the context of the discussion.
The researcher is also a study participant, so this choice is not perfect, but it seems most suited to the
interactive dimension of qualitative research. Similarly, the term key informant has a long history in
anthropology and, while rather sinister sounding, is still in common use.

Although considerably longer, more comprehensive, and (I hope) more sophisticated, this third edition is still
an introductory text designed to complement specialized works, classroom instruction and—most of all—
hands-on experience. I have included examples and citations of these works throughout to aid the reader
seeking more in-depth understanding of specific approaches and techniques.

A continuing goal of this edition is to promote greater methodological transparency as well as rigor. Paying
attention to what happens during a qualitative study—and meticulously documenting decisions and
procedures even and especially when they change—ensures that others can understand how the findings were
produced. There will always be an element of “trust me” in qualitative inquiry (compared to “trust the
methods” in quantitative studies) due to its reliance on the researcher-as-instrument. But it is incumbent on
qualitative researchers to demystify their methods as much as possible. This book is intended to help move us
farther along in that direction. When all is said and done, textbooks and coursework are necessary but never
sufficient—mastery in qualitative methods requires hands-on experience. With this in mind, I invite you on a
journey of discovery that I hope will keep you engaged and informed.

24
The Audience for This Third Edition
This book has its origins in the rapidly growing enthusiasm for qualitative methods in recent years. While
rooted in the proud tradition of social work research, the chapters take an ecumenical approach and can be
useful to a variety of practitioners and researchers in applied fields. Intended for graduate students, faculty,
and other professional researchers, this book provides a solid foundation for carrying out rigorous and relevant
research.

25
Acknowledgments

Students and colleagues have provided invaluable feedback on this text over the years, and for this I am
grateful. I also offer my deep appreciation to the study participants and research team of the New York
Recovery Study, the latter including Ben Henwood, Bikki Tran Smith, Katie-Sue Derejko, Emmy
Tiderington, Jerrold Jackson and Mimi Choy-Brown. From all of them I have learned much. Finally, I wish
to thank Kassie Graves at SAGE Publishing and the reviewers whose comments and support made this third
edition possible.

SAGE Publishing also wishes to thank the following reviewers for their kind assistance:

Barzoo Eliassi
Linnaeus University, Sweden

John D. Habif
Springfield College

Effrosyni Kokaliari
Springfield College

Lance T. Peterson
University of St. Thomas

Roberta Rehner Iversen

University of Pennsylvania

Jason H. Schaub
Buckinghamshire New University

Armin Schneider
University of Applied Sciences-Koblenz

Jessica Toft
University of St. Thomas and St. Catherine University

Bruce E. Winston
Regent University

26
1 Qualitative Methods in Context

Qualitative methods coalesced into a force to be reckoned with in the latter part of the 20th century. Their
popularity in social work, nursing, education, and other professions amplified this impact and contributed to
far greater reach in their applications. The reasons for this are many but can be summed up in the following
aphorism: “not everything that can be counted counts, and not everything that counts can be counted”
(Cameron, 1963, p. 13). Beyond providing a much-needed alternative to counting and measurement,
qualitative methods have an appeal that is unique and beguiling. With their attention to naturalistic detail and
context, qualitative reports require no decoding or statistical acumen. Yet the engaging end product of a
qualitative study belies the intense involvement and intellectual labor that came before. The work that comes
before is what this book—in its third edition—is about.

27
Differences From and Similarities With Quantitative Methods
Both quantitative and qualitative approaches rely on data collection to guide findings and conclusions.
Moreover, both are systematic even though qualitative methods entail assumptions and procedures that bear
little to no resemblance to quantitative research and statistical analyses. Though a bit simplistic, the
dichotomy, “a mile deep and an inch wide” versus “a mile wide and an inch deep,” highlights a major
distinction. Here are other distinguishing features:

Insider rather than outsider perspectives

Person-centered rather than variable-centered
Holistic rather than particularistic
Contextual rather than decontextual
Depth rather than breadth
Inductive rather than deductive (at least at the outset)

Qualitative methods favor naturalistic observation and interviewing. As such, they imply a degree of closeness
and an absence of controlled conditions that stand in contrast to the distance and control of traditional
scientific studies. Qualitative research is predicated on an “open systems” assumption where the observational
context (and the observer) is part of the study itself (Manicas & Secord, 1982). In contrast, quantitative
research favors a closed system approach in which every effort is made to neutralize the effects of the
observational context (including the observer).

Qualitative studies seek to represent the complex worlds of respondents in a holistic, on-the-ground manner.
They emphasize subjective meanings and question the existence of a single objective reality. Furthermore,
they assume dynamism, a state of flux that can only be captured via intensive engagement. Whereas the heart
of a quantitative report is its statistical findings, a qualitative report is a bricolage, pieced-together, tightly
woven whole greater than the sum of its parts.

Doing qualitative research requires an unparalleled degree of immersion by the researcher as the instrument of
data collection. As such, the qualitative researcher must be a sensitive instrument of inquiry, capable of
flexibility and on-the-spot decision making about following promising leads. The dynamic tension between
flexibility and serendipity on the one hand and rigor on the other makes qualitative research exciting and also
challenging. Most qualitative researchers feel no need to declare their work to be “science,” at least in the way
that science became defined in the mid-20th century. This was not always the case—today’s science emerged
from centuries of naturalistic observation in mathematics, astronomy, and biology.

The emphasis on local particularity and dynamic change leads some qualitative researchers to argue that the
methods cannot be taught or simulated (Hammersley, 2004). Apprenticeship is the preferred learning
modality, a trial-and-error approach in which the novice is guided by a mentor and the uniqueness of the
study is given their joint and full attention. Anthropologists have long practiced this way (many still do), but
apprenticeship is rarely possible outside of the closely supervised pursuit of a thesis or dissertation. Clearly, the

28
proliferation of textbooks and courses in qualitative methods attests to the growing demand for classroom
preparation with the understanding that ultimate mastery depends on practice (and more practice).

29
The Intellectual Life Course of Qualitative Methods

30
The Formative Years
The scope and purpose of this book do not allow an in-depth treatment of the long and complex genealogy of
qualitative methods. Nevertheless, it is useful to provide an overview of their “donor disciplines” such as
anthropology, sociology, philosophy, and linguistics (McCracken, 1988), as well as the next-generation
disciplines that have embraced and extended the methodologies known collectively as “qualitative.” The latter
include education, nursing, social work, medicine, and public health.

The longest continuous tradition of qualitative inquiry belongs to anthropology, originating with late 19th-
century ethnographies of non-Western peoples and cultures and continuing today, albeit in much different
form. The era of the Lone Ethnographer (Rosaldo, 1989), which extended until World War II, introduced
many of the defining characteristics of qualitative methods as they are known today. Interestingly, the leading
anthropologists of their day—Boas, Kluckoln, Malinowski, Lowie, Benedict, Mead—offered minimal
guidance in the how-to aspects of ethnographic fieldwork.

By the mid-20th century, the Chicago School of sociologists had begun to produce a rich and varied body of
qualitative research based on observations closer to home, including studies of medical students (Becker, Geer,
Hughes, & Strauss, 1961) and entire towns and communities (Lynd & Lynd, 1937, 1956). Based at the
University of Chicago, the sociology department pioneered an approach that became systematized as a unique
methodology, beginning with the early leadership of Robert Park, Ernest Burgess, and W. I. Thomas, and
culminating with the publication of The Discovery of Grounded Theory by Barney Glaser and Anselm Strauss
(1967). Glaser and Strauss’s vision of inductively derived midrange theories was set against two intellectual
traditions dominating 20th-century social science in the United States—on the one hand, a fascination with
the nomothetic (i.e., the grand theorizing of Sigmund Freud, Karl Marx, and B.F. Skinner), and on the other
hand, a commitment to anthropology’s trademark emphasis on localized or idiographic description. This
dialectic of nomothetic versus idiographic has been a running theme in qualitative inquiry from the earliest
years.

A number of exemplary qualitative studies of American life were produced by social scientists during this
golden era of ethnography, including Whyte’s Street Corner Society (1955), Goffman’s Asylums (1961),
Powdermaker’s Stranger and Friend (1966), Liebow’s Talley’s Corner (1967), and Stack’s All Our Kin (1974).
These works endure as classic examples of verstehen—understanding of human interactions and societal
structures via participant observation. In psychology, an early example of qualitative methods included
Festinger, Reicken, and Schacter’s classic study of a doomsday religious cult (1956).

An unprecedented move toward elucidating and codifying qualitative methods began in the 1980s, heavily
influenced by Glaser and Strauss’s grounded theory (Glaser & Strauss, 1967; Strauss & Corbin, 1994) but also
by the works of anthropologically trained researchers in the field of education such as George and Louise
Spindler, Jules Henry, and Harry Wolcott. Nursing researchers followed suit in pioneering the use of
qualitative methods in their field (Morse, 1985; Sandelowski, 1993). Social work’s embrace of qualitative
methods came somewhat later (Fraser, 1995; Gilgun, 1994; Padgett, 1998; Riessman, 2008) but has grown

31
apace (see Box 1.1). The American Psychological Association most recently turned this corner with the
formation of the Division of Quantitative and Qualitative Methods and the journal Qualitative Psychology
(Gergen, Josselson, & Freeman, 2015).

32
Box 1.1 Social Work Research and Qualitative Methods
Social work research in the United States was transformed with the founding of the Society for Social Work and Research (SSWR)
in 1995 at its inaugural conference in Arlington, Virginia. SSWR’s annual conference program reflected quantitative dominance but
qualitative methods also had a strong presence (e.g., a workshop on narrative analysis conducted by Catherine Riessman was
included in the first SSWR conference program). Though in the minority among SSWR’s members, qualitative researchers have
nonetheless grown in influence, and SSWR’s annual conference features an array of presentations drawing on diverse forms of
qualitative inquiry. SSWR gained a European counterpart in the formation of the European Social Work Research Association
(ESWRA), which held its first conference in 2011. ESWRA’s membership and conference programs leave little doubt that
qualitative methods are widely accepted in Western Europe. The international success of the journal Qualitative Social Work is
another manifestation of this phenomenon. Postmodern- and performance-inspired qualitative researchers (including a social work
interest group organized by Jane Gilgun) have an annual gathering dating to 2005 at the International Congress of Qualitative
Inquiry under the leadership of Norman Denzin (http://www.icqi.org).

Meanwhile, an impassioned call for a “science of social work” (Brekke, 2012) reinvigorated discussions of research impact within and
beyond the profession. Brekke’s lament that social work knowledge does not get translated or used and his proposed solution to
catalyze social work research into becoming more scientific have garnered attention and counter-balancing commentary (Longhofer
& Floersch, 2012). Brekke (2012) struck a nerve when he noted that the profession’s abiding commitment to helping others has
been accompanied by a troubling ambivalence about research, in effect ceding the task of identifying best practices to others.
Subsequent discussions at SSWR and in convened meetings of scholars have solidified a “science of social work” agenda.

If wider recognition and respect is contingent on an unwavering commitment to science, where do qualitative methods come in?
This question is hardly unique to social work, but what is particular to the profession is an assertion by some that qualitative methods
are optimally suited to the needs of a practice-based profession (Gilgun, 1994). In response, I have argued that the fit between social
work practice and qualitative methods is not as natural as it appears, and to think otherwise overlooks important methodological and
ethical differences (Padgett, 1998). Although opinions about this still differ, a few conclusions can be drawn at this juncture: (1)
quantitative research in social work is in no danger of losing its hegemony; (2) qualitative research in social work is thriving; and (3)
all social work researchers want the knowledge base of the profession to be respected and influential beyond its borders.

33
The Rise of Quantification and Operationalism
The life course of qualitative methods was profoundly affected by the rise of statistics and survey research
beginning in the 1930s and accelerating after the 1960s. Two converging trends made the dominance of
quantitative methods possible. First, mathematicians (Fisher, Pearson, and colleagues) began to develop
formulas for correlating numerically measured phenomena to assist in agricultural research on fertilizers and
crop yields. Second, an interest in business and marketing took root in sociology that represented a distinct
alternative to the Chicago School’s focus on immigrants and communities. Paul Lazarsfeld and colleagues at
Columbia University’s sociology department were leading proponents of this shift toward survey research,
opinion polls, and mathematically driven sampling.

The post–World War II economic boom, the Cold War arms race, and the U.S.–U.S.S.R. competition in
space led to an unprecedented degree of interest in scientific research just as statistical methods and computer
systems were being developed that could accommodate large datasets and multivariate analyses. The rise and
dominance of realism (a philosophical belief that a reality exists independent of the researcher) and
operationalism (said reality can be measured objectively) effected a dramatic change in the social and behavioral
sciences, coinciding with (and contributing to) a decline in the popularity of ethnographic and case study
approaches. Nowhere was this more apparent than in the discipline of psychology, where measurement
became the order of the day. In education, business, and the military, psychological testing and measurement
were considered the most reliable means of screening and detecting the troubled as well as the more
“intelligent” or capable. Among the social sciences, sociology, economics, and political science became heavily
quantitative by the 1970s. This shift had the side effect (whether intentional or not) of aligning some of these
disciplines with the well-funded “hard” sciences of biomedicine and technology. History and anthropology
also joined in the popularity of quantification but retained a strong allegiance to qualitative methods.

34
Epistemological Differences Among Qualitative Researchers
Positivist reasoning and quantification came under withering criticism from academic thinkers in the latter
part of the 20th century and the so-called “paradigm wars” ensued (Gage, 1989). Among the earliest critics
were sociologists Berger and Luckmann (1967) with their groundbreaking treatise on the social construction
of reality. Debates over epistemology, that is, philosophical discussions of how we know what we know,
consumed many in the academy. Building toward an antipositivist epistemology, postmodern critics found
common cause with French intellectuals including Jacques Derrida and Michel Foucault. A number of
philosophers and social scientists in Europe and the United States joined this burgeoning movement
(Harding, 1987; Rabinow & Sullivan, 1979). In anthropology, Geertz (1973, 1988) and Clifford and Marcus
(1986) inspired a profound reexamination of ethnography in light of previous assumptions of naïve realism.

Following the writings of Foucault (1980), critics charged that scientific inquiry was a vehicle for reinforcing
power differentials favoring a technocratic and corporatist elite. The traditional dichotomy between “hard”
and “soft” sciences (the latter used in a pejorative manner) was but one manifestation of how scientific inquiry
(sometimes derided as “scientism”) rested on a foundation of realism. Since objectivity is a chimera, critics
argued, the distance between the researcher and the researched was an arbitrary and self-serving premise. Self-
consciousness and reflexivity were presented as necessary correctives to traditional notions of distance and
power or control in research.

Beginning in the late 1980s, disciplinary boundaries became blurred as some qualitative researchers turned to
the humanities—philosophy, literature, and the arts. Two major turning points came with the publication of
Naturalistic Inquiry by Lincoln and Guba (1985) and the first of several editions of the Handbook of
Qualitative Methods edited by Denzin and Lincoln (1994). The Handbook’s appearance heralded a new era of
constructivism and critical theories; its goal was to shape the destiny of this growing phenomenon known as
“qualitative methods.” As noted in the introduction, “there are no objective observations, only observations
situated in the worlds of the observer and the observed” (Denzin & Lincoln, 1994, p. 12). Venturing further
down this road in a later edition of the Handbook, Denzin and Lincoln soundly rejected standards of
quantitative methods in favor of “verisimilitude, emotionality, an ethic of caring, political praxis, multi-voiced
texts” (2011, p. 9).

Interestingly, the paradigm wars were heating up at roughly the same time that qualitative methods were
surging in popularity in nursing and social work. For students in these professions, learning about qualitative
methods meant being drawn into epistemological camps in which one was urged to declare allegiance to one
camp or another (Staller, 2012). The alternative—“epistemological unconsciousness”—was viewed as a default
endorsement of positivism (Steinmetz, 2005, p. 12). The labels given to this resistance—postmodern,
antifoundational, poststructural, constructionist, interpretivist—grew in number and nuance.

Constructionism—a belief that human phenomena are socially constructed rather than objectively “real”—
proved to be a liberating force for many researchers (Denzin & Lincoln, 2011; Holstein & Gubrium, 2007). It
led to reexamination and critique of what is meant by “race,” “sexuality,” and “mental illness” among many

35
other social “facts.” Exploring the ways that such concepts are invented and reified challenges assumptions of
an immutable reality and value-free research. It is not a big step to move from constructionism to critical
theories including feminist, Marxist, race, and queer theories (Harding, 1987; Hesse-Biber, 2006; Ladson-
Billings, 2000; Madison, 2005; Olesen, 2000). All were united in pointing out the hidden (and not-so-
hidden) subtexts of knowledge produced by Western science.

A related term, constructivism, was coined to underscore the importance of subjectivity and reflexivity. In
explaining the “constructivist turn” in her interpretation of grounded theory, Charmaz (2014) argued that the
benefits of constructionist research had been undermined by the lack of adequate recognition of the researcher’s
role in the social constructions being described. Charmaz’s work signaled a further departure from what she
and others called the objectivist roots of grounded theory. In recent years, social constructionism and social
constructivism have come to overlap in meaning such that the two are frequently used interchangeably
without further explanation.

Postmodern criticism and epistemological self-awareness have had an undeniable impact in many academic
quarters. Yet it would be overreach to say that their influence has been wholly transformative since most
qualitative studies published in journals have been and remain epistemologically agnostic (Alvesson &
Karreman, 2011). Some qualitative researchers have criticized the postmodernist turn (Seale, 2004), decrying
its cycles of “advocacy and exegesis” (Flaherty, 2002, p. 509). For researchers outside of the comfortable
remove of academia (or laboring in its untenured realms), constructivist epistemologies and postmodern
criticism come with palpable risks as well as intellectual rewards.

36
The Pragmatic Middle Ground
Acknowledging the benefits of critical thinking (and theories) and taking a more utilitarian view of methods is
the essence of pragmatism in qualitative research. A uniquely American phenomenon developed by John
Dewey, Charles Peirce, and Jane Addams (among others), pragmatic philosophy was a reaction to
metaphysical arguments on the nature of truth and reality (Cherryholmes, 1992; Menand, 2001; Rorty, 1998).
Rather than take a stand on such philosophical conundrums, pragmatists accept the fallibility of knowledge
development, elevating utility over ideology or philosophy. Thus, one can be comfortable with the notion that
there are occasions when realist claims should be made (“genocide in Nazi Germany”), when a presumed
reality practically cries out to be deconstructed (the concept of “race”), and when multiple subjective meanings
produce a broader understanding of something (“chronic pain”). Put another way, all concepts are human
inventions, but some are more socially contrived and consequential than others.

The Chicago School’s symbolic interactionism and pragmatic philosophy were closely intertwined (John
Dewey and George Herbert Mead were friends and colleagues in the philosophy department at the University
of Chicago). Thus, human experiences—beliefs as well as actions—were cast as socially situated and inquiries
into such experiences subject to varying interpretations. Several qualitative methodologists (Becker, 1996;
Creswell, 2013; Morgan, 2014; Patton, 2002) have gone on record favoring pragmatism. Such an espousal is
not a simple matter of choosing the path of least resistance. Rather, pragmatic philosophy in the Dewey
tradition denies the need to choose between postpositivism and constructivism as incompatible paradigms
with opposing views of the existence and nature of reality.

A point of contention comes into focus when considering constructivists’ assertions that their work is more
closely aligned with social justice concerns (most clearly articulated in the Denzin and Lincoln Handbooks).
Dewey and his colleague, Jane Addams, hardly shied away from social action on behalf of the poor and
disenfranchised. In doing so, they were following a central premise of pragmatism tying beliefs to action
(Morgan, 2014). More recent applications of pragmatism to feminism and other critical theories have
produced a form of “reflective pragmatism” (Gergen, Josselson & Freeman, 2015, p. 2) in which social values
are integrated along with attention to rigor. Pragmatism has been further bolstered by the surge of interest in
mixed methods (Morgan, 2014; Tashakkori & Teddlie, 2010), their popularity a de facto rejection of the
incompatibility thesis of constructivism (more on this topic in Chapter 10). In the end, pragmatism is about
more choices, not fewer.

37
Evidence-Based Practice and Implementation Science
Evidence-based practice (EBP) has been a powerful movement in the health and social services (Gibbs &
Gambrill, 2002; Morse, 2006). Rooted in biomedicine, EBP presents a challenge to the practicing professions
to offer evidence-based interventions and abandon those found to be ineffective or harmful. Determinations
of effectiveness are made using a hierarchy of evidence that places randomized experiments at the top followed
by quasi-experimental then nonexperimental studies.

In line with the EBP movement, the Cochrane and Campbell Collaborations (the former dedicated to health
and the latter to social welfare, education, and criminal justice) have generated numerous reports from meta-
analyses and metasyntheses showing the effectiveness (or lack thereof) of interventions ranging from welfare-
to-work programs to antipsychotic medications (see www.cochrane.org and www.campbellcollaboration.org).
It is perhaps not surprising that Cochrane reviews have multiplied while the Campbell initiative has been
slowed by the complexities of social welfare research as well as the scarcity of experimental trials in the
literature. Even when available and amenable to meta-analysis, the evidence can be problematic given the
dynamic, often-changing conditions of social work, human services, educational, and criminal justice systems.

The narrow definitions of evidence and rigor attending EBP determinations typically preclude studies that
could explain the how and why of interventions (not just the whether). Although efforts have been made at
Cochrane to include qualitative findings through systematic reviews (Noyes, Popay, Pearson, Hannes &
Booth, 2008), they have been slow going given the lack of standardized reporting and conformity in such
studies.

The EBP movement clearly has its limitations, although its underlying premise is difficult to repudiate. Who,
if anyone, would want to use a surgeon who ignores the latest research findings? At the same time, acceptance
of what EBP has to offer need not lead to devaluing other forms of knowledge. A step in the right direction
has been taken by some leading quantitative researchers who have gone on record favoring qualitative methods
as a “touchstone of reality” in community-based interventions (Hohmann & Shear, 2002, p. 205).

In recent years, the field of implementation science has become prominent, an outgrowth of the EBP movement
(Proctor et al., 2007). In addition to the U.S. National Institutes of Health, organizations such as the World
Bank, Global Fund, and World Health Organization have embraced implementation science. Developed to
put the spotlight on real-world conditions surrounding dissemination, implementation science addresses key
questions raised by failed attempts at spreading EBPs. The rationale for doing so is clear: millions of dollars
are wasted each year on sophisticated but misinformed randomized trials that do not work in the venues
where they were intended to ultimately succeed. Outside of the controlled conditions of a laboratory-like
environment, what appears to be a rock-solid intervention can crumble quickly if the translation is faulty.

Implementation science offers a systematic conceptual framework for understanding and mitigating this
process of translation (Damschroder et al., 2009). In doing so, it introduces a necessary realization of the
benefits of multiple methods (Palinkas et al., 2011). Qualitative methods in particular allow examination of

38
the “backstage” happenings and unexpected consequences taking place in the programs and clinics where
innovations succeed or fail. As shown in the journal Implementation Science, these innovations range from
family violence screening to diabetes self-care.

Implementation science does not, however, portend major methodological change—it still adopts the prestige
hierarchy placing randomized experiments at the top. Nor does it leave much room to consider structural and
macrolevel barriers as opposed to locally tangible and measurable factors. Yet acknowledging the importance
of qualitative and mixed methods (Palinkas et al., 2011) in implementation research is a step in the right
direction.

39
Theoretical and Conceptual Frameworks in Qualitative Inquiry

40
Some Are a Better Fit Than Others
The relationship between qualitative research and theory is subject to varied opinions (Alvesson & Karreman,
2011; Anfara & Mertz, 2006). On the one hand, allowing one or more theories to drive the inquiry deprives a
study of what qualitative methods do best (the production of theory being a key contribution of grounded
theory). On the other hand, an atheoretical qualitative study—one without any prior conceptualization—risks
being irrelevant and marginalized. The choices in this regard are many. In the current climate of multiple
epistemologies and paradigmatic influences, specific methods bring their own disciplinary influences and
conceptual frameworks (more on this topic in Chapter 2).

It is helpful to distinguish among several versions of what is meant by theory in the social and behavioral
sciences. This variety in the meaning of theory is due to differing degrees of explanatory ambition, conceptual
abstraction, and openness to multiple interpretations. These differing meanings include (1) grand theories
having a sweeping scope and high level of abstraction (e.g., Freudian or Marxist theory); (2) somewhat less
ambitious theories in psychology and sociology, of which some are more amenable to being operationalized
and tested than others (an example of the former is Bandura’s Self-Efficacy Theory and of the latter is
Bowlby’s Attachment Theory); (3) critical theories (feminist, race, queer, etc.) that address societal
inequalities; and (4) theories that operate as an “open system” and illuminate more than explicate (e.g., Mead
and Blumer’s Symbolic Interaction Theory or Bronfenbrenner’s Social Ecology Theory). Falling short of full
status as a “theory” are (5) conceptual frameworks that offer organizing principles and evocative concepts for
research without being strongly predictive (e.g., the Health Belief Model and the Andersen and Newman
model of service utilization) and (6) inductively derived “mid-range” theories that have been the foundation of
grounded theory methodology.

Perhaps not surprisingly, Versions 1 and 2 are least suited for qualitative studies to the extent that they are
determining rather than orienting in their intent and use. Psychological and biomedical theories present a
distinct challenge because they are focused on identifying and solving problems of an intrapsychic or physical
nature. As such, their underlying propositions are less open to nonpathologizing interpretations. Critical
theories (Version 3) have been adopted by qualitative researchers but are less about methodology than
ideology and advocacy. In contrast, Versions 4 and 5 are a closer fit for qualitative research given their
openness to inductive reasoning. Version 6 is only possible with qualitative research.

41
Analytic Induction and Abduction: Nondeductive Thinking
Inductive reasoning—collecting empirical data without having to test a theory—is the sine qua non of
qualitative methods. When the researcher is focused on inductive theory—building from the data (admittedly
not the goal of all qualitative studies), analytic induction offers an iterative method: a straightforward process of
choosing a case and developing a theory for it, then choosing another case to see if the causal theory fits, and
so on. When a case is encountered that does not fit (a “negative case”), the researcher is confronted with the
decision to keep trying or to refine the theory in some way. This corresponds to Karl Popper’s thesis that one
may surmise all swans are white until that one black swan shows up. Owing much to the writing of Znaniecki
(1934), analytic induction was a direct response to the rise of statistics and their capacity to enable deduction
through operationalization (Bernard & Ryan, 2010). Like its famous progeny of grounded theory (Glaser &
Strauss, 1967), analytic induction offered a powerful counterpoint to “grand theories” as well as to statistical
tests. The method spun off a quantitative variant when Charles Ragin (1987) introduced qualitative
comparative analysis (QCA) where testing alternative theoretical propositions followed Boolean algebra or
logic.

Cressey (1953) provided a classic example of analytic induction in his study of imprisoned embezzlers.
Beginning with his first interview and through subsequent interviews with the men, he developed and tested
his theories for why they stole from their employers. His early explanations were rebutted by negative cases,
beginning with “they did not think it was illegal” (they knew it was illegal) and moving on to “they needed the
money because of debt” (some did not need the money but embezzled anyway). However, the last hypothesis
became refined to fit the circumstances of all 133 embezzlers Cressey ultimately interviewed. That is, the men
did not necessarily need the money, but they were experiencing a personal crisis of some kind and believed
that the money would somehow help them through it. They maintained a sense of rectitude by reasoning that
they were using the funds for personal betterment rather than absconding with them (Cressey, 1953).

Then there are the lesser-known but no less ardent proponents of abduction who criticize the inductive
approach that spawned grounded theory. With origins attributed to pragmatist philosopher Charles Peirce
(1934), abduction (Latin for “to lead away”) is rooted in the assumption that hypotheses are best developed in
the face of counterintuitive evidence or what Agar (2006) calls “rich points.” Timmermans and Tavory (2012)
argue for abduction by introducing their own “rich point”: the surprising absence of viable theories arising
from a plethora of grounded theory studies. Although this claim is a bit contentious, several abductive
alternatives have emerged, including the extended case method (Burawoy, 1998), analytic ethnography (Snow,
Morrill, & Anderson, 2003) and situational analysis (Clarke, 2005).

Timmermans and Tavory (2012) describe abduction as neither atheoretical (as presumed with induction) nor
theory driven and deductive. Abduction begins with the observation of a perplexing problem or conundrum,
then proceeds to reference existing theories or explanations to develop a new and better situated hypothesis
that circles back into what is known to test its explanatory value. In contrast to analytic induction, abductive
thinking at its best depends upon a well-informed tenacious researcher comfortable with ambiguity and with
being on a quest. It is not a case-by-case testing approach.

42
Abduction has proven valuable in business and technology where being bewildered or surprised is not
tolerated very well (Agar, 2006). An example of abductive problem solving can be found in Vaughan’s (1996)
insightful analyses of the cascade of poor decisions leading to the U.S. space shuttle Challenger explosion.
Agar (2006) offers a simpler example of abduction in relating the true story of a new but unused medical clinic
built near a Latino community in Texas. Baffled that residents did not flock to the new clinic, a researcher
spoke to the residents and found that the only public transportation (a bus) ran directly to the old clinic but
nowhere near the new one.

43
The Importance of Concepts
Qualitative researchers rely on a variety of theories and conceptual frameworks to mine relevant concepts
(known as “sensitizing concepts” in grounded theory). Here again, some are a better fit than others. The
concepts of coping and self-esteem, for example, can be borrowed from psychology, but their strong association
with hypothetico-deductive theories could produce confusion when applied to subjective experiences. Even
less applicable are terms that imply clinical judgments or unconscious processes, for example, denial or
transference from psychodynamic theory. As mentioned earlier, open systems theories such as symbolic
interaction and social ecology are a good fit for use in qualitative research.

Concepts less deeply embedded in theoretical frameworks or professional jargon are more likely to be useful.
Social support is an example as are stigma and self-determination. As Barney Glaser (2002) aptly noted, concepts
have staying power and the capacity for representation and evocation that transcends literal description.
Considering theories as lenses—as alternative ways of refracting meaning from data—produces a kaleidoscope
of possibilities in line with constructionist thinking. The challenge is to look for fit, not through precise
alignment but through achieving different ways of understanding. The contribution of a theory or a concept
has to be established (it is never guaranteed), but without conceptual frameworks coming before and from
data analysis, the study’s contribution is severely diminished.

44
The Place and Timing of Theories in Qualitative Studies
Starting from the premise that one can draw on extant theories and concepts in a qualitative study, a few
questions naturally arise:

What (if any) theoretical ideas and concepts are to be used?

When and how do they inform the study?
How are they incorporated into the study?

Answering these questions depends on the study’s epistemological foundation as well as its approach and
methodology. If epistemology is viewed as occurring upstream and methods downstream, theories and
conceptual frameworks operate in between. Figure 1.1 depicts this flow from the abstract to the concrete with
examples given for each phase. The arrows pointing to the right in the upper two boxes indicate the
continuing influence of epistemology and of theories or concepts; the feedback loop arrow at the bottom
connects methods and data collection iteratively.

The “what” question builds on this discussion with the additional caveat that qualitative researchers may
simultaneously draw on several theoretical frameworks and concepts as lenses through which the study’s data
and ideas are refracted (Marshall & Rossman, 2010). This is a good reason for adopting a multidisciplinary
perspective, because openness to ideas from a variety of sources lends freshness and creativity to a qualitative
study.

How might this work? A study of family caregivers of schizophrenia patients might draw on concepts from
social exchange theory (such as reciprocity), from the research literature (such as burden and stigma), and
from the researcher’s practice orientation (such as resilience and recovery). These represent a place to start but
hopefully not to finish, their survival dependent on whether they earn their way into the findings (Charmaz,
2006).

Figure 1.1 The Foundations and Processes of a Qualitative Study

45
The “when” question about theories refers to timing, that is, if theories influence the study from the beginning
or are held in abeyance until the data analysis and interpretation phase. The latter approach is often used in
phenomenological research, which favors immersion and new insights. Qualitative researchers following a
postmodern approach are likely to reject existing interpretive frameworks as diminishing the importance of
voice and insider perspectives, at times substituting an ideological framework of critical thinking. Most
qualitative approaches involve theoretical ideas and concepts drawn from the review of the literature. These
ideas and concepts may remain during the analyses and new ones may be incorporated as well.

The “how” question is the most challenging and least discussed. Theoretical influence usually takes the form
of concepts that sensitize and orient the researcher during data collection. However, some theories such as
symbolic interactionism consist of a web of ideas and framings that influence the study in a more subtle way.
Even if adhering to a purer form of induction, theorizing in some form takes place in the data analysis stage.
This is where interpretation kicks in, taking the study beyond data reduction and description.

46
Reasons for Doing Qualitative Research
It might seem unnecessary to rationalize the need for qualitative research in a book featuring these methods,
but there are reasons that stand out as well as reasons not to use them. Below are a few scenarios where
qualitative methods are an obvious choice.

1. You are exploring a topic about which little is known—especially from the “inside” perspective. This approach is
the hallmark of qualitative methods. There are many fascinating research topics, for example, risky sex among
men “on the down low” who proclaim themselves heterosexual, women who are impregnated by rape and
decide to keep the child, parents who adopt severely disabled children, family members who participate in
assisted suicide. What is important is that too little is known about them and an in-depth understanding is
sought.

2. You are pursuing a topic of sensitivity and emotional depth. Social workers and other human service
professionals routinely encounter human crises and dilemmas that require empathy and understanding. These
experiences provide a wellspring of ideas for research. For researchers interested in behaviors considered taboo
or stigmatized, qualitative methods may be the only plausible approach. Qualitative studies of cocaine dealers,
gang members, and sex workers portray the lives of individuals who are not likely to cooperate with the usual
forms of survey research. Moreover, studies of sensitive topics need not be confined to the fringes of polite
society—one can “study up” as well as “study down” (e.g., community leaders after a natural disaster, business
executives with a gambling addiction, doctors who abuse prescription drugs). Of course, members of an elite
population are often the hardest to study, capable of using their power to limit access (Hertz & Imber, 1995).

3. You wish to capture the “lived experience” from the perspectives of those who live it and create meaning from it.
When researchers seek verstehen (deep understanding), they pursue studies that are emic, that is, focused on
the insider point of view rather than etic, or the outsider’s perspective. Examples might include studies of
persons with physical disabilities, survivors of partner violence, or first responders in disaster relief.

4. You wish to get inside the “black box” of practice, programs, and interventions. Perhaps not surprisingly given the
push for accountability in social services, program evaluations have become heavily focused on quantitative
outcomes. Yet qualitative methods have a secure place in evaluation research and implementation science.
They are a natural fit with formative evaluations given their capacity to identify unforeseen effects of a new
program that may hamper (or pave the way to) its implementation. Likewise, qualitative methods in process
evaluation shed light on how (not whether) a program succeeds or fails. So much of professional practice plays
out in messy, unbounded ways that do not lend themselves to standardized measurement. Qualitative studies
do not yield “hard” outcomes, but their naturalism and agility can produce a description that emerges
organically from real-world settings.

5. You are a quantitative researcher who has reached an impasse in explaining or understanding. It is striking how
often unanswered questions emerge during quantitative studies that call for qualitative research. My earlier
quantitative research on ethnic differences in mental health services help-seeking frequently led me to fall

47
back on a “cultural” explanation calling for more in-depth examinations of how members of ethnic groups
perceive mental illness and the service delivery system (Padgett, Patrick, Burns, & Schlesinger, 1994). The
insurance claims database we used was of no use for such a purpose.

6. You are seeking to merge advocacy with research. Action-oriented participatory research is devoted to countering
the effects of oppression and social injustice (Fals-Borda & Rahman, 1991; Freire, 1973; Reason & Bradbury,
2008). Rooted in the efforts of community activists, feminists, and liberation movements of the 1960s and
1970s, action research seeks to fuse knowledge building with advocacy. Although both qualitative and
quantitative methods can be used, the central premises of action research are closely aligned to the
relationships common to qualitative research.

7. You wish to study complex social processes. This final reason, added by my colleague Julianne Oktay (2012), has
loomed large in recent years. The 21st century has brought increasing recognition of “wicked problems” that
in turn highlight the limits of existing research methods (Mertens, 2015). Research on climate change, causes
of war, and new infectious diseases has brought renewed interest in ecological and systems perspectives that
are not amenable to the reductionism of statistical measurements and calculations (Diez-Roux, 1998).
Similarly, individuals exist within complex ever-changing social systems, and their lives unfold in ways that
cannot be reduced to numbers (Singer, Ryff, Carr, & Magee, 1998). “Thinning the descriptions of individual
lives” to variables that standardize and generalize risks losing nuance, complexity, and interconnections
(Singer et al., 1998, p. 6). Enter qualitative methods.

In summary, there are many sound reasons to do qualitative research—some or all of the aforementioned
scenarios may give rise to a particular study. There are also reasons not to pursue qualitative research.
Foremost among these is that the topic of interest is better served by quantitative designs such as experiments
or surveys. Second, anyone seeking qualitative research as “the easy way” should be forewarned—the intensive
labor and immersion required are reason enough to think twice.

48
Desirable Qualities and Skills in the Qualitative Researcher-as-Instrument
A qualitative study’s success depends heavily on the researcher’s personal qualities as well as intellectual
capacity. The relative absence of structure allows wide latitude—to reach creative heights as well as the depths
of intellectual paralysis or personal bias. A few qualities that an individual can have or cultivate make the
conduct of qualitative research more successful. These include flexibility, reflexivity, and an ability to multitask
in an iterative, nonlinear way.

Flexibility is a state of mind and of behavior well-suited to the unpredictable, ever-changing landscape of
qualitative inquiry. This happens on a number of fronts. Respondents may suddenly refuse to cooperate, put
off being interviewed, or not show up at all. They may divulge shocking information, make sexual overtures,
or suddenly turn the questioning around to put the interviewer on the spot. Data analysis may (and often
does) lead to new directions and new study participants. One’s favorite ideas or theories may not be borne out
and, as a result, may have to be cast aside. The strength and success of qualitative research lie in the
researcher’s ability to go with the flow rather than try to control it.

Reflexivity, the ability to critically examine one’s self, is a central preoccupation since the qualitative researcher
is inextricably part of the study. To paraphrase Michael Agar (1980), the problem is not whether the
qualitative researcher is biased; the problem is what kinds of biases exist and how can their impact be
managed. Examining one’s biases begins before the study begins and continues thereafter.

Finally, a qualitative study is guaranteed to produce vast amounts of raw data awaiting management and
analysis from its earliest stages. The ability to multitask on a number of levels (i.e., to simultaneously collect
and analyze data, keep track of what is happening via memos, and remain open to new insights and the bigger
picture) must be present and nurtured in the researcher-as-instrument.

In addition to personal capabilities, certain skills are essential to the qualitative research enterprise. Among
these are observation and interpersonal communication. When teaching qualitative methods, I ask my social
work students to carry out an exercise in participant observation. They must go to a public place of their
choice (a park, subway station, street fair, playground, etc.), observe the action for one hour, and write up field
notes describing what they have seen. What a departure this is for them! Trained to engage clients and focus
on problem resolution, they must be passive observers of all they can take in. The open-ended nature of
qualitative observation can be awkward and even painful for individuals who are more comfortable with the
“filters” of clinical and personal theories and the authority to guide what happens. The interpersonal skills of
empathy and sensitivity, so important in clinical practice, are put to somewhat different ends in qualitative
research. Rather than foster engagement with clients for treatment purposes, these skills enable listening as
part of the pursuit of knowledge and understanding. This requires a degree of humility and subordination of
self that takes some getting used to.

Two additional skills needed in qualitative research are the interrelated abilities to think conceptually and
write well. The need to think abstractly and create new perspectives lends interpretive meaning to data. This is

49
not done in a vacuum. The researcher’s “interpretive repertoire” includes the paradigmatic, methodological,
and theoretical knowledge that guides but also constrains the work (Alvesson & Karreman, 2011). Moreover,
formulating ideas and developing concepts and theories are interwoven with the ability to write. The act of
writing is a means of thinking through and framing the issues.

A well-developed sense of humor helps enormously in qualitative research, particularly the ability to laugh at
oneself. One’s vulnerability and inexperience when starting out almost guarantee that there will be mistakes;
some of them will be funny in the eyes of others. (Anthropologists invariably have stories of abject humiliation
and jokes made at their expense.)

Finally, there is the capacity for collaboration. Although individuals can and still do carry out qualitative
studies with little outside assistance, the scope and sophistication of research makes teamwork and
collaboration increasingly the norm. Multidisciplinary participation is especially welcome in a qualitative study
in which the richness of insight is enhanced by differing perspectives.

50
Studying the Familiar Versus the Unfamiliar
There are so many potential topics of study that the choices can seem endless. The physical setting can be
nearby or halfway around the world. The topic can arise from a number of places, including personal and
professional interests as well as intellectual curiosity about some aspect of the human condition as yet
unknown. This distinction between studying the familiar versus the unfamiliar (Ely, Anzul, Friedman,
Garner, & Steinmetz, 1991) has far-reaching implications.

The temptation to stick to the familiar can be powerful. Many a valuable study has had its origins in personal
biography. Among doctoral students I have known, studies of Alzheimer’s patients, interracial couples, and
runaway adolescents were motivated by personal concerns. Professional interests and accessibility often
converge. Hence, teachers carry out studies in schools, nurses conduct research with patients, and
psychologists study their undergraduate students. For social workers, the familiar is typically social service
agencies and their clients. Studying practice settings is quite natural for many professions because the pursuit
of knowledge has an ultimate goal of improving practice.

There are two main advantages to studying the familiar:

1. Easier development of rapport: With the familiar, the path to acceptance and cooperation is usually
smoother and comfort levels are higher. This is no small benefit, as gaining access to a research site and
to respondents can be very demanding for the qualitative researcher.
2. A head start in knowledge about the topic: Studying the familiar presumes prior knowledge about the
subject, either through personal or professional experience.

There are obvious disadvantages. Foremost among these is the risk of being too close. Consider an example of
an agency supervisor interested in studying her own program. What happens if/when the staff members find it
difficult to accept her in this new (and considerably humbler) role as a qualitative researcher? Those who
cooperate might give prepared answers, worried that they might be judged harshly. The remedy for this is
straightforward but far from easy. “Defamiliarizing,” or viewing the research setting as exotic rather than
known, presumes an ability to confront one’s pre-existing notions (Alvesson & Karreman, 2011).

Fascination with the unfamiliar, a longstanding motivation in qualitative research (especially ethnography),
offers advantages. Starting with the vantage point of an outsider means having the distance needed to discover
tacit rules and norms. In this instance, role confusion is less problematic because one enters the field with the
identity of a researcher. Of course, studying the unfamiliar can be risky. Just as knowing too much creates
blinders, knowing too little can lead to stereotyping and “filling in the blanks” prematurely. In addition,
accessibility and acceptance are more demanding when entering unfamiliar territory. One must negotiate with
formal gatekeepers as well as individual respondents to enlist their cooperation and trust. Reversing the
process—making the unfamiliar familiar—requires patience and persistence (Ely et al., 1991).

51
52
Values and Social Responsibility in Qualitative Research
Qualitative methods have an inherent appeal to practitioner-researchers because they are rapport-driven and
do not presume value-free inquiry. It is no accident that a growing number of researchers in social work are
attracted to qualitative methods as compatible with their political and personal values (Padgett, 2009). Many a
study participant has remarked on the quasi-therapeutic effects of in-depth interviews and the gratification of
being listened to in an engaging empathic exchange.

Beyond its salutary effects on participants, qualitative research is compatible with social justice and
emancipatory values (although this is not inherent to the methods). One means of expressing these values is to
stress human agency and a strengths perspective as a counterbalance to the pathological approach borrowed
from medicine and underlying much of clinical practice as it has historically developed (Saleebey, 2005). In
qualitative research, respect for human agency is manifested in the ways that human actions are observed and
interpreted, in the questions asked, and in the interpretation and presentation of the findings. One does not
manufacture strengths if they do not appear to exist. Similarly, the temptation should be resisted to portray
study participants solely as victims to draw attention to their plight.

How does a socially conscious researcher maintain this balancing act? One way to do this is by distinguishing
between personal strengths and larger structural forces. Consider, for example, a hypothetical study of
mentally disabled persons confronted with cutbacks in their government benefits. When writing up the study
(assuming the data go there), we take note of their resilience, that is, how they seek assistance from family and
friends amidst tremendous financial stress. Thus, the message of the study is one of perseverance in trying
circumstances. This could be illustrated using case histories of individuals living in fear of eviction, their
mental fragility worsening, and their informal resources depleted. In this way, the report can become part of
public advocacy without sacrificing the nuance and detail that the methods provide.

What happens if the findings directly conflict with a goal of social advocacy? It is highly unlikely that a study’s
results would directly contribute to or worsen misfortune in people’s lives. However, the temptation to distort
or mislead, even for noble purposes, should be resisted.

53
Introducing the New York Services Study (NYSS) and the New York
Recovery Study (NYRS)
Being fortunate enough to receive two all-qualitative grants from the National Institute of Mental Health
(NIMH) in 2004 and in 2010, my colleagues and I embarked upon longitudinal studies of the lives of
homeless adults after enrolling in two very different supportive housing programs. These men and women—
all of whom had diagnoses of serious mental illness and histories of substance abuse and homelessness—were
generous enough to share their stories with us. We also interviewed (with their permission) the program case
managers who oversaw their care. The first study (the NYSS) focused on service engagement and relied
exclusively on in-depth interviews. This second study (the NYRS) followed participants (and their program
case managers) over 18 months. In addition to in-depth interviews, the NYRS had an expanded qualitative
methods portfolio, including ethnographic site visits, “shadowing” interviews and photo-elicitation interviews
(the latter consisted of giving study participants cameras to document their lives visually as well as verbally).
Specific examples from the NYSS and the NYRS will be used in this book to illustrate various facets of a
qualitative study.

When filtered through the content of this chapter, the NYRS was rooted in a pragmatic epistemology that
accommodated the empiricism of multifaceted data collection without foregoing the possibilities of
interpretation rooted in social constructionism. The NYRS was informed by the following theoretical
frameworks: Bronfenbrenner’s ecological perspective (1979), Freire’s writings on empowerment (1973), and
the capabilities approach of Amartya Sen (1999) and Martha Nussbaum (2000).

The place and timing of theories in the NYRS was early and often. In the first place, their presence was
essential to the success of the grant proposal as the National Institutes of Health expects funded research to be
situated within existing conceptual frameworks and theories (even if not explicitly testing them). Secondly,
these theories shaped how we viewed formerly homeless adults’ lives and their adjustment to supportive
housing. This meant examining their lives as embedded within environmental systems—their families,
friends, program staff, the neighborhood, the city and beyond. Finally, these theories were applied during
analyses as provisional schemas that may or may not map onto the data. New theories and concepts can be
(and were) brought into the analysis phase. Likewise, earlier theories and ideas were set aside—such is the
responsibility of staying close to the data.

Among the various reasons for doing qualitative research described earlier in this chapter, the NYRS aligned
with them all. Thus, we sought to capture the insider perspectives of clients and staff working in supportive
housing programs, we knew that we would be addressing sensitive and emotionally laden topics (trauma,
addiction, mental illness), and we hoped to comprehend and convey the lived experience of formerly homeless
adults struggling with their own trajectories of recovery from and in mental illness. The “black box” of
program operations associated with supportive housing was also a focus although not tied to evaluation goals.
The raison d’etre for the NYRS was to shed light on aspects of clients’ lives that were not measured or
measurable in previous research.

54
The final reason—the pursuit of advocacy—was not foregrounded in the study (a strategic decision given the
funder’s priorities). Nevertheless, we used the language of empowerment unapologetically (Padgett &
Henwood, 2009). One of the supportive housing programs being studied used a “housing first” approach
based on the right to housing and consumer choice regarding compliance with medications and substance use
(Tsemberis, Gulcur, & Nakae, 2004). The comparison programs used the traditional approach in which
clients were expected to comply with rules including medication compliance and sobriety and did not have
their own apartment. Part of the “science” of the NYRS lay in its rejection of a foregone conclusion that the
“housing first” participants would manifest superior trajectories of recovery. The investigative team (including
this writer) was inclined toward such an outcome (and the research evidence was firmly behind such a
supposition), but we remained open to unanticipated possibilities.

In revisiting the desirable qualities of a qualitative researcher, the NYRS team was carefully assembled to
maximize the capacity for openness, empathy, flexibility, tenacity, and humor as well as attention to the
minutiae of study protocols. Mindful that we were all-too-human, the team met weekly to debrief and
practice collective reflexivity, checking and balancing as much as possible the inevitable tendency to prejudge,
distort, or overlook. The team was deliberately comprised of members who were studying the familiar—an
anthropologist, a social work faculty member, and graduate students in social work and public health who had
prior experience working with the homeless and with persons with serious mental illness. Making the familiar
unfamiliar was not easy, but having different types of experiences and differing perspectives kept things fresh
and less prone to groupthink. A strong commitment to social justice united members of the team, but
members were hardly in lockstep on how that should be infused into the study and its findings.

55
Summary and Concluding Thoughts
Qualitative researchers are a committed and occasionally rambunctious group. Beginning with a rich historic
background synonymous with ethnography (and, later, grounded theory), qualitative methods became
consolidated as a vibrant alternative to quantitative methods after the 1970s. They also became more diverse
and multidisciplinary. Modern influences ranging from phenomenology to postmodern criticism brought
variety in epistemology and methods. By the 1990s, the various nonquantitative methods that became known
collectively as “qualitative” came into full flower—with phenomenological and narrative approaches joining
ethnography, grounded theory, and case studies. Although not explicitly qualitative, action-oriented
approaches proved to be friendly terrain for many qualitative researchers.

The flourishing of qualitative methods took place despite (and because of) the dominance of quantification in
Western science. Made possible by the development of statistical analyses and the post–World War II arms
and space races, this dominance has eroded somewhat in recent years as the limits of operationalism have
become apparent (not everything that counts can be counted). The rise of evidence-based practice has
renewed interest in aggregate (and aggregated) findings, but it has also raised further questions about the
nature of evidence and the methods used to determine its existence.

This chapter introduced key topics that cut across the diverse landscape of qualitative inquiry, including the
role of theories and concepts, reasons for using qualitative methods, the advantages and disadvantages of
studying the familiar versus the unfamiliar, and social responsibility in research. Addressing each of these
orients the reader to the complex and discretionary aspects of qualitative methods—all of which will be in
evidence in the chapters to come.

56
Exercises
In the classroom, break into small work groups. (This can also be done individually.) Choose a research topic
of interest:

1. Discuss the various epistemological positions described in this chapter. How might your topic be framed
in terms of these paradigms (postpositivist, constructivist, critical)?
2. What advantages do qualitative methods bring to the study?
3. Is the topic familiar to one or more group members? How is familiarity both an advantage and
disadvantage in carrying out the study? What are some ways to make the “familiar become unfamiliar?
4. Are there theoretical frameworks or concepts that could be brought to bear on the chosen topic of
study? If yes, discuss how this might happen.
5. What (if any) are the ways that social responsibility is reflected in the chosen topic?

57
Additional Readings
Berg, B. L., & Lune, H. (2011). Qualitative research methods for the social sciences (8th ed.). Westwood, NJ:
Pearson.

Bodgan, R., & Biklen, S. J. (2006). Qualitative research for education (5th ed.). Westwood, NJ: Pearson.

Bourgeault, I., Dingwall, R., & DeVries, R. (Eds.). (2013). The Sage handbook of qualitative methods in health
research. Thousand Oaks, CA: Sage.

Carey, M. (2012). Qualitative research skills for social work. London, UK: Ashgate.

Crabtree, B. F., & Miller, W. L. (1999). Doing qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Creswell, J. W. (2012). Qualitative inquiry and research design (3rd ed.). Thousand Oaks, CA: Sage.

Denzin, N. L., & Lincoln, Y. S. (Eds.). (2005). Handbook of qualitative research. Thousand Oaks, CA: Sage.

Flick, U. (Ed.). (2013). The Sage handbook of qualitative data analysis. London, UK: Sage.

Flick, U. (2014). An introduction to qualitative research (5th ed.). London, UK: Sage.

Floersch, J., & Longhofer, J. (2012). Qualitative methods for practice research. New York, NY: Oxford.

Hesse-Biber, S., & Leavy, P. (2010). The practice of qualitative research. Thousand Oaks, CA: Sage.

Marshall, C., & Rossman, G. B. (2015). Designing qualitative research (6th ed.). Thousand Oaks, CA: Sage.

Merriam, S. (2002). Qualitative research in practice: Examples for discussion and analysis. New York, NY: J.
Wiley & Sons.

Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis (2nd ed.). Thousand Oaks, CA: Sage.

Miles, M. B., Huberman, A. M., & Saldana, J. (2013). Qualitative data analysis: A methods sourcebook.
Thousand Oaks, CA: Sage.

Morse, J. M. (2012). Qualitative health research: Creating a new discipline. San Francisco, CA: Left Coast
Press.

Oktay, J. (2012). Grounded theory. New York, NY: Oxford.

Padgett, D. K. (Ed.). (2004a). The qualitative research experience. Belmont, CA: Thomson.

Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

Rossman, G. B., & Rallis, S. F. (2011). Learning in the field (3rd ed.). Thousand Oaks, CA: Sage.

Savin-Baden, M., & Howell-Major, C. (2013). Qualitative research: The essential guide to theory and practice.

58
New York, NY: Routledge.

Silverman, D. (2010). Doing qualitative research: A practical handbook (4th ed.). London, UK: Sage.

Tolman, D. L., & Brydon-Miller, M. (Eds.). (2001). From subjects to subjectivities: A handbook of interpretive
and participatory methods. New York: New York University Press.

59
Journals That Feature Qualitative Studies and Methods
American Anthropologist
Culture, Medicine & Psychiatry
Discourse Studies
Families & Society
Field Methods
Forum: Qualitative Social Research (FQSR)
Human Organization
International Journal of Qualitative Methods
International Journal of Social Research Methodology
Journal of Contemporary Ethnography
Journal of Mixed Methods Research
Journal of Phenomenological Psychology
Journal of the Society for Social Work and Research (JSSWR)
Medical Anthropology Quarterly
Narrative Inquiry
Narrative Works
Qualitative Health Research
Qualitative Inquiry
Qualitative Psychology
Qualitative Research
Qualitative Social Work
Qualitative Sociology
Social Science & Medicine
The Qualitative Report

60
Websites for Qualitative Methods
http://www.nsf.gov/pubs/2004/nsf04219/start.htm (excellent proceedings from workshop on qualitative
methods at the National Science Foundation)
http://www.uofaweb.ualberta.ca/iiqm/Conferences.cfm (comprehensive site from the University of
Alberta in Canada, who sponsors the leading international conference annually)
http://www.nova.edu/ssss/QR (online journal “The Qualitative Report”)
http://www.quarc.de (German–English online resource)
http://qualitative-research.net (German–English–Spanish site with online journal)
http://ejournals.library.ualberta.ca/index.php/IJQM/index (International Journal of Qualitative
Methods)
http://www.researchtalk.com (training/workshop company)
http://www.methodspace.com (forum for researchers)

61
Selected Monographs

62
Early Classics
Becker, H., Geer, B., Hughes, E., & Strauss, A. (1961). Boys in white: Student culture in medical school.
Chicago, IL University of Chicago Press.

Bosk, C. L. (1979). Forgive and remember: Managing medical failure. Chicago, IL: University of Chicago Press.

Estroff, S. (1981). Making it crazy. Berkeley: University of California Press.

Gans, H. (1962). The urban villagers: Group and class in the life of Italian-Americans. New York, NY: Free
Press.

Goffman, E. (1961). Asylums: Essays on the social situation of mental patients and other inmates. Garden City,
NY: Basic Books.

Hochschild, A. (with Machung, A.). (1989). The second shift: Inside the two job marriage. New York, NY:
Avon.

Liebow, E. (1967). Talley’s corner: A study of Negro street corner men. Boston, MA: Little, Brown.

Lynd, R. S., & Lynd, H. M. (1956). Middletown: A study in modern American culture. New York, NY:
Harcourt Brace.

Myerhoff, B. (1978). Number our days: A triumph of continuity and culture among Jewish old people in an urban
ghetto. New York, NY: Simon & Schuster.

Painter, N. I. (1979). The narrative of Hosea Hudson: His life as a Negro communist in the south. Cambridge,
MA: Harvard University Press.

Powdermaker, H. (1966). Stranger and friend: The way of an anthropologist. New York, NY: Norton.

Scheper-Hughes, N. (1981). Saints, scholars and schizophrenics: Mental illness in rural Ireland. Berkeley:
University of California Press.

Stack, C. (1974). All our kin. Strategies for survival in a black community. New York, NY: Harper Colophon.

Whyte, W. F. (1955). Street corner society (2nd ed.). Chicago, IL: University of Chicago Press.

63
Later Examples
Abrams, L., & Anderson-Nathe, B. (2012). Compassionate confinement: A year in the life of Unit C. New
Brunswick, NJ: Rutgers University Press.

Anderson, E. (1999). Code of the street: Decency, violence, and the moral life of the inner city. New York, NY: W.
W. Norton and Company.

Biehl, J. (2013). Vita: Life in a zone of social abandonment. Berkeley: University of California Press.

Bourgois, P. (2009). Righteous dope fiend. Berkeley: University of California Press.

Duneier, M. (1999). Sidewalk. New York, NY: Farrar, Straus & Giroux.

Floersch, J. (2002). Meds, money and manners. New York, NY: Columbia University Press.

Goffman, A. (2014). On the run: Fugitive life in an American City. New York, NY: Picador.

Holmes, S. (2013). Fresh fruit, broken bodies: Migrant farmworkers in the United States. Berkeley: University of
California Press.

Iversen, R., & Armstrong, A.L. (2006). Jobs aren’t enough: Toward a new economic mobility for low-income
families. Philadelphia, PA: Temple University Press.

Karp, D. A. (1996). Speaking of sadness: Depression, disconnection and the meaning of illness. New York, NY:
Oxford University Press.

Klinenberg, E. (2002). Heat wave: A social autopsy of disaster in Chicago. Chicago, IL: University of Chicago
Press.

Liebow, E. (1993). Tell them who I am: The lives of homeless women. New York, NY: Penguin.

Martin, E. (2007). Bipolar expeditions: Mania and depression in America. Princeton, NJ: Princeton University
Press.

Moller, W. D. (2004). Dancing with broken bones: Portraits of death and dying among inner-city poor. New York,
NY: Oxford University Press.

Newman, K., Fox, C., Roth, W., & Mehta, J. (2005). Rampage: The social roots of school shootings. New York,
NY: Basic Books.

Oktay, J. (2005). Breast cancer: Daughters tell their stories. Binghamton, NY: Haworth Press.

Rhodes, L. (2004). Total confinement: Madness and reason in the maximum security prison. Berkeley: University
of California Press.

64
Young, A. A. (2004). The minds of marginalized black men. Princeton, NJ: Princeton University Press.

65
2 Choosing the Right Qualitative Approach(es)

Qualitative research represents different things to different people. This chapter is devoted to six of the most
commonly used methodological approaches: ethnography, grounded theory, case studies, narrative,
phenomenological, and action-oriented research. Because there are detailed how-to books on each of these,
this overview is designed primarily to assist in deciding which approach to use—the reader is urged to read
further once an approach is selected (a list of additional readings is provided at the end of this chapter).
Singling out each approach for in-depth description belies the fact that qualitative researchers frequently mix
and match approaches within the confines of a single study (a topic covered later in this chapter).

66
Six Primary Approaches in Qualitative Research

67
Ethnography
Ethnographic research, the most senior of the “elders” in the qualitative family, has been enshrined as a
method, a theoretical orientation, and a philosophical paradigm (Tedlock, 2000). Although its popularity has
ebbed and flowed over the years, ethnography has maintained its central position as the quintessential
qualitative method. Its reliance on direct observation and emic (or insider) perspectives sets a high standard for
commitment that stands in contrast to the etic (or outsider) perspective assumed by many researchers (Goetz
& LeCompte, 1984).

In addition to requiring skills in gaining rapport and engaging in intense and ongoing observation,
ethnography implies an attitude or stance. Specifically, it means that one adopts a holistic perspective, viewing
all aspects of the phenomenon under study as parts of an interrelated whole. Ethnography also embraces
cultural relativism, a perspective holding that cultures must be understood on their own terms, not judged by
the beliefs and values of other, more powerful cultures. Although intolerable if taken to the extreme (e.g.,
considering the genocide in Rwanda to be a manifestation of tribal enmities that should not be judged by
outsiders), cultural relativism has value as a challenge to ethnocentrism, or the denigration of cultures other
than one’s own. Despite the trademark approach of participant observation, anthropologists have for a long
time incorporated measures and statistical analyses in their work (e.g., changes in caloric intake or group
differences in social networks).

Doing ethnography means focusing on a cultural system with identifiable features (Lofland & Snow, 2005).
The boundaries may be physical such as the walls of a hospital or the perimeters of a neighborhood, or they
may be defined by shared identities (e.g., gang members, a professional football team, runaway adolescents).
Ethnographic inquiry means operating on several levels simultaneously to infer the tacit rules of the culture or
subculture from the myriad of actions and interactions being witnessed. The staple of ethnographic data is the
field note, a meticulous record of what is (or was) observed. Field notes are rarely scintillating reading, but
there is no substitute and the sooner done, the better (memory fades quickly). Chapter 5 offers more details
on ethnographic methods and Box 2.1 provides an example of an ethnographic study.

Ethnography as a methodology is nonprescriptive and thus offers considerable latitude, but this freedom
comes at a price. Ample time and labor are needed to get it right (and what is right is not always agreed-upon
or apparent). Nevertheless, ethnographic methods have enjoyed increasing popularity among qualitative
researchers seeking to capitalize on their unique contribution to understanding (Bernard & Gravlee, 2014).

68
Box 2.1 Compassionate Confinement: Ethnography in a Juvenile
Detention Center
Laura Abrams and her colleague, Ben Anderson-Nathe, acted on a shared interest in juvenile corrections to conduct an ethnographic
study at Wildwood House (a pseudonym). Gaining entrée is rarely without snags, but criminal justice facilities are notoriously
resistant. Having the acquaintance of the facility director who, in turn, supported research proved to be pivotal for Abrams and
Anderson-Nathe.

In their book entitled Compassionate Confinement (2013), the authors describe the study and its inevitable twists and turns. Having
access did not translate into ease in gaining assent from the young people or from the parents and guardians they sought out during
visiting hours. Forbidden from offering incentives, Abrams and Anderson-Nathe relied on prolonged engagement and trust building
to encourage residents to speak freely and privately (the facility provided office space for one-on-one interviews). Taking turns with
weekly field observation over 16 months produced copious field notes entered into two columns—the left-hand column a long
running descriptive commentary on what was being observed and the right-hand column reserved for thoughts, memos, journaling,
and reactions.

Participation was a constant negotiation—should they join in a pick-up basketball game? Eat together with the boys? The staff? Sit
in on therapy groups? As collaborators, they needed to work out shared responsibilities and ensure quality control over data
collection and during interactions with staff and directors. Being open to surprises and setbacks and sharing in information-rich
meetings and interviews paid off in relatively smooth entrée as well as in ending the study successfully (Abrams & Anderson-Nathe,
2013).

69
Grounded Theory
Grounded theory (GT) has emerged as the most widely used approach in qualitative research since its debut
in the late 1960s (Glaser & Strauss, 1967). Closely aligned with the Chicago School of sociology, GT sparked
broad interest fueled by the growing popularity of in-depth interviewing in the 1970s and beyond. GT’s
procedural instructions helped to demystify qualitative methods and make them accessible to researchers
across a wide range of disciplines including social work (Longhofer, Floersch & Hoy, 2013; Oktay, 2012).
What also set GT apart was its rejection of grand theories in favor of “small t” theories with less ambition but
more authenticity or “groundedness.” This was a revelation for its time, since the 20th century was an era of
soaring and ambitious theories (Freud, Marx, Sartre, Foucault) of which few were amenable to measurement
and testing.

GT has evolved significantly over the years, surviving a dispute between Glaser and Strauss (the latter of
whom was joined by Juliet Corbin in subsequent works). The early parting of ways led Strauss into academia
(University of California at San Francisco) and Glaser into establishing the Grounded Theory Institute.
Differences between the two pivoted on Strauss’s later published works, which offered more in-depth
instruction in the methods (Corbin & Strauss, 2007). In contrast, Glaser (2002) emphasized less structure and
more emphasis on theoretical conceptualization. Despite such differences, GT has grown enormously in
popularity over the years and its accessibility has undoubtedly led to the wider acceptance of qualitative
methods.

The central goal of a GT study is to explain a process or action. Thus, it does not have the freeze-frame
approach common in phenomenological analysis and traditional ethnography. The object of a GT study is
often a life transition—Glaser and Strauss’s (1965) classic study of dying patients formed the cornerstone for
the methodology. The dynamism inherent in GT comes from individual interviews with persons undergoing
(or having undergone) a transition or life change. This can be fairly routine (e.g., going to college) or highly
unusual (e.g., winning the lottery). Sample sizes in GT are moderate (20 to 30 is about right, but samples can
be smaller or larger).

While Chapter 6 will feature much more information on GT data analysis, a summary can be offered here.
GT entails inductive coding from the data, memo writing to document analytic decisions, and weaving in
theoretical ideas and concepts without permitting them to drive or constrain the study’s emergent findings. In
an elegant inversion of the theory-driven deduction common to quantitative research, GT has made the
pursuit of midrange theories a respectable, even desirable outcome of qualitative research.

Cycling between data collection and analysis, GT begins with open coding of interview transcripts. The process
of coding may use sensitizing concepts drawn from the literature, extant theories, and previous research, but its
primary goal is inductive. Coding proceeds to axial and selective phases, gradually creating a parsimonious
conceptual framework featuring categories. Along the way, the researcher employs constant comparative analysis
to examine contrasts across respondents, situations, and settings. The concept of saturation is used to guide
when data collection and analysis comes to an end. Saturation is achieved when new participants (or codes or

70
themes) are redundant and no longer contributing new information.

There are challenges to conducting a GT study, for example, recognizing when saturation has occurred or
when theoretical sampling should be undertaken. The features that make GT attractive—relatively codified
procedures and guidance—can also constrain creativity. Yet the concerns many qualitative researchers have
about GT are not with the methods per se but with how they are applied. A kind of “GT lite” has emerged in
the research literature in which codes and themes are the endpoint, and the study does not carry through to
completion the development of categories and their properties and a fully articulated grounded theory. Box
2.2 features a comprehensive grounded theory study that endures as an exemplar.

71
Box 2.2 Surviving Childhood Sexual Abuse: An Exemplary Grounded
Theory Study
In their grounded theory study of 11 women who had survived childhood sexual abuse, Susan Morrow and Mary Lee Smith (1995)
give the lie to assertions that “small n” qualitative studies cannot be rigorous and productive of valuable knowledge. A diverse group
of participants, the women were interviewed individually and a subgroup of seven became a 10-week ongoing focus group that also
contributed to the study as coresearchers. To give substance to their claims of evidentiary adequacy (Erickson, 1986), Morrow and
Smith reported that the data consisted of “over 220 hours of audio- and videotapes which documented more than 165 hours of
interviews, 24 hours of group sessions, and 25 hours of follow-up interactions” over a 16-month period (1995, p. 25). Supplemented
by 16 hours of audiotaped field notes and reflections, the data amounted to over 2,000 pages of transcriptions, field notes, and
documents.

Meticulously following grounded theory procedures, Morrow and Smith identified over 160 different coping strategies adopted by
the women. Further analysis and interpretation yielded categories that were interconnected in explaining how the coping strategies
and survival emerged, intervening phenomena, and the consequences of the strategies when used. Each category was fully described
and illustrated with quotes from the women. In addition to delineating cognitive and psychological responses to sexual abuse,
Morrow and Smith note the resilience and forbearance of the women as survivors rather than victims.

72
Case Study Analysis
Although case studies are not exclusively qualitative, they have a long and honorable history in qualitative
research (Feagin, Orum, & Sjoberg, 1991). The term case study can be confusing as it refers to both method
and product. As studies of “bounded systems of action” (Snow & Anderson, 1991, p. 152), case studies draw
on the ability of the qualitative researcher to extract depth and meaning in context. A specific location—a
pediatric ward, an immigrant neighborhood or a nursing home—can be the focus of a case study. A “case”
may also be a pivotal event, for example, a landmark Supreme Court decision or the closing of a military base.
Case studies often play an important role in program evaluation (Greene, 2000). A study of an exemplary
hospice program, for example, can offer insights into best practices.

Noteworthy events can provide an opportunity to explore historical and social changes. Eric Klinenberg’s Heat
Wave (2002) used media reports, documents, and interviews to provide a “social autopsy” case study of the
disastrous effects of Chicago’s 1994 heat wave on the elderly poor. Regardless of its subject matter, the case
study draws on multiple perspectives and data sources to produce contextually rich and meaningful
interpretation. In this regard, it is important to distinguish case studies in qualitative research from their
counterparts in clinical training and business education. Clinical cases—a commonly used pedagogical tool for
training students in psychiatry, psychology, nursing, and social work—illustrate the application of clinical
theories to individual cases. Similarly, business schools use detailed cases of entrepreneurial successes or
failures as learning tools. In qualitative research, the case study is a method of inquiry for knowledge
development that necessitates systematic processes of data collection and analysis (Donmoyer, 1990).

How does one carry out a qualitative case study? There are a number of useful how-to guides (Gerring, 2007;
Hancock & Algozzine, 2006; Merriam, 1998; Mills, Derepos, & Wiebe, 2010; Stake, 2005; Yin, 2013). Of
these, the most frequently cited authorities are Robert Yin and Robert Stake. Yin and Stake employ somewhat
different approaches (Yin associated with positivist epistemology and Stake tending to adhere to constructivist
interpretations). Although the terminology varies, the instruction in the methodology by Yin and Stake has
much in common.

According to Stake (1995), case studies fall into three types: intrinsic, instrumental, and collective. An intrinsic
case study is suitable for an exploratory topic or one that is rare or vivifying enough to merit description and
analysis. An instrumental case study has a larger purpose, that is, to exemplify or demonstrate something, to
develop theory, or to evaluate and critique. In other words, the case is a means to an end rather than an end in
itself. The collective case study corresponds to multiple case analyses and is more likely to aggregate
instrumental than intrinsic cases (Stake, 2005).

The first step for the researcher is one of bounding the case in space and time. If the subject matter is a
discrete event, the boundaries are set and recognizable. If the topic of interest can be tagged to a number of
cases, for example, welfare agencies adjusting to a dramatic change in policy, it is incumbent on the researcher
to make clear which agency (or agencies) will be studied and why. Circling back to Stake’s typology, the
discrete and rare case is more likely to be intrinsic; the nondiscrete and nonrare case is more likely to be

73
instrumental. In a study of a single case, it may be intrinsic or instrumental, but it should be compelling
enough to warrant investigation.

Once the case(s) have been identified, data collection begins. Although true in other types of qualitative
research, the use of diverse sources of data is critical in case studies, for example, interviews, observation, and
archival documents. Using these data, the researcher puts together a thorough description of the case, its
origins, its development, and any change over time. If the case has component parts, these are described and
their relationship explained. The larger context is also described as it influences the case—social, cultural,
political, and economic conditions.

If there is more than one case, the researcher takes the extra step of identifying common themes or patterns
across the cases. Here, the challenge is one of aggregating across cases while maintaining the distinctive nature
of each case (Campbell & Arens, 1998; Ragin, 1987; Stake, 2005; Yin, 2013). This can be a messy process,
and it is invariably iterative, that is, reading and rereading notes and interviews to bring fresh insights and
ideas. Keeping the case intact can produce different types of findings, for example, cases are often clustered
into types that are mutually exclusive and exhaustive. Typology development done this way stands in contrast
to codes drawn from interviews. Challenges of case study analysis include selecting the right case(s), gathering
sufficient data, and analyzing multiple cases without sacrificing the holistic qualities of each.

Case studies resemble ethnography in their emphasis on holism and use of multiple data sources (and Box 2.3
shows both may be used in tandem), but the researcher’s motives for conducting the study usually differ.
Ethnography is dedicated to cultural analysis; its vitality comes from “thick description” of beliefs and
behaviors. For example, an ethnographer interested in a breakaway religious sect might use participant
observation, interviews, and documents (letters, religious tracts, etc.) to understand it as a subculture with its
own norms and beliefs governing members’ behavior. The same group could be the subject of a case study
analysis, an intrinsic case study if its inner workings are worth describing to a larger audience or an
instrumental case study exemplifying psychological control over others or as part of a larger societal move
toward fringe religious beliefs. When the object(s) of inquiry require holism over disaggregation, case study
analysis is most likely the route to take.

74
Box 2.3 Introducing an Evidence-Based Program Into a Hostile
Setting: A Case Study Using Ethnographic Methods
Felton (2003) used ethnographic methods to carry out a case study that portrayed the tumultuous introduction of an innovative
evidence-based practice known as Housing First (HF) in a suburban New York county. Adopting a constructivist approach featuring
local actors’ perspectives, she conducted interviews with a variety of stakeholders affected by this abrupt change in policy, including
county officials overseeing social services, mental health providers, homeless shelter operators, and HF staff.

Opposition to HF was vociferous and rooted in a fear that this new approach—which placed homeless persons with serious mental
illness directly into their own apartments without demonstrating “housing readiness”—was irresponsible and dangerous. Despite
previous experiences resulting in no such dire consequences, the opposition began with letters to the newspaper and built into an
organized town hall meeting in which county officials and representatives of the HF agency were accused of improper contract
negotiations as well as flagrant disregard of community norms.

Using documents as well as interviews—news stories, letters to the editor, service contracts, county budget reports, and local housing
reports—Felton delved deep into her case and wove together an account of how innovation took place in volatile circumstances that
eventually settled peacefully into a “new normal.” The opposition was neither a grassroots community movement nor was it
orchestrated by business leaders as might be expected. It was, instead, the creation of local shelter providers fearful of losing ground
to a new model that might supplant them. In time, the HF program gained a foothold and proved effective in housing the homeless,
and yet this did not result in the shelter’s demise. In retrospect, Felton’s case study is both intrinsic and instrumental, that is, it has
depth and texture coming from many interviews with stakeholders, but it also transcends the particulars of time and place to offer
insights into larger issues attending innovations in human services.

75
Narrative Approaches
Narrative approaches (NA) have tremendous intuitive appeal given their emphasis on the power of the spoken
word (Andrews, 2014; Czarniawska, 2004; Mattingly, 1998; Mishler, 1986; Polkinghorne, 1988; Riessman,
2008; Wells, 2011). Indeed, their popularity and widespread invocation have extended the utility of narratives
into therapeutic and self-help domains in which clients are asked to “re-story” their lives (White & Epston,
1990). Our interest is with NA as a diverse set of research methods focusing on how something is said as well
as what is said.

Rooted in literature, history, and sociolinguistics, narrative approaches start from the premise that speaking
and writing are forms of meaning-making (Frank, 2010). As such, NA methods work best when studying one
(or a few) individuals and the topic lends itself to “narrativizing” of some kind (Clandinin & Connelly, 2004).
Life histories, biographies, folk tales, and psychotherapy sessions are fertile ground for studies in “narrative
knowing” (Polkinghorne, 1988).)

The popularity of narratives has prompted greater scrutiny into their role in relation to personal identity and
the creation of the self (Snow & Anderson, 1987). The interrelatedness of the self and narrative can be
illustrated by the ways that personal stories evolve to enhance their personal meaning and the teller’s
psychological well-being. Trauma survivors, for example, often adjust their narratives over time to enhance
comfort and meaning (e.g., “it must have happened for a reason”). Some researchers express concerns about
conflating narrative and self, stating that this results in linguistic reductionism of complex inner
“conversations” and cognitions (Smith & Sparkes, 2006). Spotlighting texts can crowd out larger contexts as
well as the existence of nonnarrated “selves.” For example, a study of elderly South Africans that left out the
historical context of apartheid would be a missed opportunity.

Moving to context shifts the standpoint from micro-examinations of text to the dialogic interaction that
produced the narrative and from there to broader environmental influences (Frank, 2012; Riessman, 2008).
Although the dialogue of interest is usually between the researcher and participant, it may occur among
participants and be witnessed or recorded by the researcher. Interactions between social workers and their
clients or between parents and children often contain important narratives or conversational content.

When the focus is on dialogue rather then narrativizing, the NA may be conversation analysis (CA) or discourse
analysis (DA). With roots in sociology and ethnomethodology (Gubrium & Holstein, 2000; Sacks &
Garfinkle, 1970), CA examines sequencing, turn taking, “holding the floor,” interruption, and other aspects of
conversation that reveal how social roles and identities are manifested during talk (Farnell & Graham, 2000).
Audiotaped transcriptions of conversations can be analyzed with CA to offer clues to how interpersonal
communication both shapes and reflects social interaction (Sidnell & Stivers, 2013). Angell and Bolden
(2015) used CA to analyze transcripts of 36 conversations between psychiatrists and patients with serious
mental illness to understand how medications were prescribed and rationalized.

Discourse analysis (DA) emerged as a technique for identifying the social meanings reflected in talk and text

76
(Fairclough, 2003; Gee, 2005). Meaning can be ascertained from a variety of indices, including choice of
words and idioms, speaking rhythm and cadence, inflection, intonation, gestures, and nonverbal utterances
(groans, sighs, laughter, etc.). Foucauldian discourse analysis, which draws on Foucault’s critiques of
hegemonic power, tends to operate at a more abstract level than the everyday discourses of interest to most
qualitative researchers. Standing in contrast to such analyses are the best-selling books by sociolinguist
Deborah Tannen (1990, 2006) in which she explores how men and women “just don’t understand” one
another and analyzes the volatile communications between mothers and daughters. Both CA and DA depend
on transcripts of conversations, but their analytic power is enhanced by recording nonverbal sounds as well as
behavior.

There is an inherent understanding that interpretation—whether done by the researcher alone or in

collaboration with the participant—is essential to all NA. By their nature, narratives are retrospective accounts
not intended to be factual reproductions. This of courses begs the question of what and how. The rise of
postmodernist thinking with its focus on textual exegesis was a natural fit with narrative analysis, although far
from the only influence.

What are the data analytic methods of NA? Labov’s microstructural classification and labeling method has
attained popularity as both a stand-alone technique and starting point. Riessman (2008) notes that the
integrity of the narrative and its constituent parts is a necessary yet rarely sufficient step in data analysis.
Recognizing that the narrative is located in larger contexts—in time, place, and audience—leads one to
broader analytic goals. The story is kept intact, but—similar to case study analysis—one may “theorize from
the case rather than from component themes (categories) across cases” (Riessman, 2008, p. 53). Pederson’s
(2013) study of job loss narratives used thematic analysis to create a typology of identities describing
individual responses to unemployment.

Studies of narratives and speech fill an important niche within qualitative methods, highlighting the
importance of language at the interpersonal level but also within larger social and political contexts. With an
emphasis on the ways that language shapes (not simply mirrors) reality, they bring a natural fit with
constructivist interpretations. Gender, race, and other social inequalities permeate the way we speak to one
another and how we narrativize our life experiences.

77
Phenomenological Analysis
Phenomenological analysis (PA) owes much to the early 20th-century philosophical writings of Edmund
Husserl and Martin Heidegger and to later developmental work in psychology, education, and nursing (Lopez
& Willis, 2004). Indeed, a proper introduction to PA requires engaging with phenomenological philosophy
and its underlying assumptions. Foremost among the latter is the awareness that reality is a product of one’s
consciousness, and the traditional boundaries between subject and object are an artifact of the “empirical turn”
in philosophy and science in the 20th century.

One carries out a phenomenological study by first identifying the phenomenon of interest. This is not a trivial
question as some phenomena are more worthy of study than others. Most topics in PA have resonance as
aspects of the human condition that run deep, for example, cancer treatment, adopting a child, or grief over
the loss of a spouse. PA puts the focus on deeper meanings achieved by prolonged immersion, that is,
capturing the lived experience. Analyses of phenomenological interview data are conducted to find the
“essence” or common themes in their experiences. Phenomenological findings explore not only what
participants experience but also the situations and conditions surrounding those experiences.

The above description contains the common elements of what have become distinct yet overlapping
approaches in PA. Following van Manen (2002) in the field of education, hermeneutic phenomenology involves
the analysis of texts describing the phenomenon or experience (these may be interview transcripts or other
written materials). The researcher is the interpreter or mediator of the varied meanings that such texts yield.
Perhaps not surprising given phenomenology’s concern with cognition, the discipline of psychology has
several leading proponents including Moustakas (1994), Giorgi (1985), and Smith (1996).

Moustakas follows a transcendental phenomenology approach that foregrounds the need for “bracketing,” or
sidelining preconceptions about what is real (what Husserl terms epoche). Thus, the researcher is not entirely
absent from the study but makes a concentrated effort to be self-aware enough not to intrude on the essential
aspects of participant’s accounts. Also heavily influenced by Husserl are Giorgi’s descriptive phenomenology
(1985, 2009) and a body of phenomenological research that emerged under Giorgi’s leadership at Duquesne
University.

Psychology also gave rise to interpretive phenomenological analysis (IPA). As described by Smith (1996), IPA is
closely aligned with health psychology but also owes a debt to symbolic interaction theory given the latter is
concerned with meaning making by individuals (Blumer, 1969). Similar to van Manen’s (1990) framing of the
researcher’s role as involved rather than “bracketed,” Smith refers to a “double hermeneutic” in which
participants are trying to make sense of their experiences, and the researcher is trying to make sense of their
making sense (Smith & Osborn, 2009, p. 53). The possibilities of using IPA are many, and research topics
naturally arise: How do women experience menopause as a life transition? What is the lived experience of
Asperger’s syndrome? How do sex offenders adjust to life after prison?

The field of nursing has also contributed to PA with work by Munhall (2012) and a guide to hermeneutic PA

78
(Cohen, Kahn, & Steeves, 2000). In social work research, PA is less common, but examples are growing in
number. Armour and colleagues used hermeneutic PA to study individuals who lost loved ones to homicide
(Armour, Rivaux, & Bell, 2009), and IPA has been applied to research topics central to social work such as
foster care (Houston & Mullan-Jensen, 2011)

Methods of data collection and analysis bear a marked resemblance across the different PA approaches. PA
interviews with around 6 to 10 participants (this number can be smaller or larger) begin with broad, open-
ended questioning to ensure the rapport and openness necessary to access their lived experiences. Multiple
interviews with each participant are needed to achieve needed depth. Reading (and rereading) interview
transcripts and any other data, the researcher flags key statements, quotes, and contexts and begins the
painstaking process of examining these across participants’ accounts. These are typically clustered into
composite themes that form the architecture of the findings. Moustakas (1994) urges researchers to include a
section in the report documenting their perspectives and experiences in conducting the study.

PA clearly occupies an important niche in the qualitative methods family, as no other approach seeks to
ensure that readers feel as if they have “walked a mile in the shoes” of study participants. At the same time, it
shares with other qualitative methods considerable latitude in how a particular study is carried out and
evaluated as successful. Subtypes of PA tend to be more philosophically than methodologically distinct, and
researchers uncomfortable with exploring the philosophical foundations of PA risk depriving it of its original
strengths.

79
Action and Community-Engaged Research
Action research (AR), traceable to the seminal work of Kurt Lewin (1946), has roots in pragmatism (Tandon,
1996) as well as 1960s liberation movements. Closely linked to participatory action research (PAR) and
community-based participatory research (CBPR), AR shares with PAR and CBPR fundamental commitments
to community empowerment and egalitarian partnerships (Reason & Bradbury-Huang, 2013; Stringer, 2013).

These approaches fit loosely under the broader category of community-engaged research (CER). CER refers
to studies that are dependent on a level of community involvement that may or may not entail an action or
participatory component. Virtually any study conducted in a community (however defined) might be CER,
but the implication here is that some degree of active engagement with the community is conducted.
Moreover, that engagement is respectful of local norms and sensitive to the impact of the research on the
community.

All CBPR and PAR is CER, but not all CER has an action or participatory component. Here is a
hypothetical example of CER that is not CBPR or PAR: researchers are interested in conducting household
interviews and ethnographic observation in a neighborhood beset by local gangs. In this instance, communal
action is not a safe option and could split the community into factions. Instead, the researchers rely upon a
low-intensity, friendly but unintrusive presence in the neighborhood. Their observation (of gang tags or
graffiti on buildings, of gathering spots for young adults, of the presence of children and the elderly (signaling
a sense of safety) is paired with household interviews regarding residents’ strategies for protecting their
families. In this example, organizing for action might come later on but is inappropriate given safety concerns
as well as a lack of understanding of the context of gang activity.

The popularity of CBPR can be traced to post-1960s movements advocating community empowerment in
general (Fals-Borda, 1998; Freire, 1973) and power sharing in research in particular (Foster-Fishman,
Berkowitz, Lounsbury, Jacobson, & Allen, 2001; Nelson, Ochocka, Griffin, & Lord, 1998). Impetus has also
come from pragmatic concerns surrounding the necessary (but complicated) move away from academic-based,
controlled trials to “real-world” interventions in communities (Hohman & Shear, 2002). The problems
attending such a change, often dismissed as “noise” by quantitative researchers, include low rates of
recruitment and high rates of study attrition. Feasibility and relevance—clinical, cultural, and social—suffer
when there is little or no buy-in from a community.

Closely linked to public health but also gaining ground in other fields (Danso, 2015; Jones & Wells, 2007),
CPBR is a natural fit for qualitative researchers in social work and other practicing professions. CBPR can be
seen as embracing “three Ps.” In other words, it is a perspective that infuses a study from start to finish; it
connotes a partnership of equals among researchers and community participants; it requires active participation
by all parties. CBPR partnerships tend to work best when all parties are willing to commit time and resources.
The potential for methodological trade-offs and compromises is considered worth the benefits in the form of
community-led improvements in health and well-being (Cornwall & Jewkes, 1995; Israel, Eng, Schulz, &
Parker, 2005). Although not inherently qualitative or quantitative in methodology, PAR and CPBR rarely

80
operate without qualitative methods, either as stand-alone or as part of mixed methods. All involve a degree of
local immersion and engagement that fits well with low-threshold qualitative methods and a “nothing about
us, without us” ethos (Nelson et al., 1998).

Challenges come from the need for time to build and nurture the partnership, from the compromises needed
to reach consensus and make progress toward shared goals, and from different, sometimes conflicting,
priorities among stakeholders. Because time and resources are almost always limited, a premium is put on
abbreviated and focused methods that can yield findings in a short turnaround time and having wide-reaching
impact. Clearly, not all research topics point to CER, PAR, or CBPR, nor are these approaches easy in any
sense of the word. Yet their contribution to applied and practice-oriented research is unique, and the mutual
respect they engender is a welcome change from the usual one-sided power relationships of researchers and
study participants.

The popularity of CBPR has spawned online resources for researchers (see the list at the end of this chapter).
Moreover, definitions of community have stretched beyond geographic boundaries to include groups based
upon shared identities, workplaces, and aspirations. In one such study led by Rogerio Pinto, the community is
comprised of HIV providers in New York City, a diverse group of nonprofit organizations with services
ranging from medical clinics to housing to mental health care. With an emphasis on understanding
interagency collaboration, a community collaborative board oversees the project and shares in decision making
at each stage of the study (Pinto, Spector, & Valera, 2011). Box 2.4 offers a cautionary tale of what happens
when researchers do not know enough about the “community” they seek to help.

81
Box 2.4 The Price of Not Knowing the Community: A Failed Trial of
an HIV Prophylactic Product
An example of the importance of knowing one’s community (and the costly consequences of not knowing) comes from the results of
a $94 million trial of an anti-HIV prophylactic administered in vaginal gel and tablet form (Marrozzo et al., 2015). Published in the
New England Journal of Medicine, the trial was conducted with 5,209 African women in Uganda, Zimbabwe and South Africa and
concluded, “None of the drug regimens we evaluated reduced the rates of HIV acquisition in an intention-to-treat analysis.
Adherence to study drugs was low” (Marrozzo et al., 2015, p. 509).

What happened? The authors point to deception on the part of the women since reported adherence was high, but actual use of the
products was low. They recommend that future researchers adopt “measures of adherence that do not rely solely on self reporting
and that are not easily manipulated by participants, such as real-time biologic monitoring of drug levels” (Marrazzo et al., 2015, p.
516). However, qualitative interviews with 102 South African women who participated in the trial revealed that adherence was not a
simple matter (van der Straten et al., 2014). The requirement of daily use was probably the greatest misstep that caused the trial to
fail, as it was burdensome as well as ill suited to the nature of sexual activity with male partners who are migrant workers (Susser,
2015). Moreover, adherence had been found to be higher in studies where the women were asked to apply the gel before and after
sex. Taking the time to understand the women’s lives would have saved considerable expense and finger-pointing.

In noting this failure and the Ebola crisis that occurred in Africa later the same year, Susser (2015) writes, “Bottom-up research
design may improve results, but this takes time, costs money and disrupts accepted hierarchies. Because funders and donors may not
recognize the need to build in the costs of community engagement, studies are more likely to focus on pharmaceutical methods than
on strong investment in local participation.”

82
The Six Approaches Revisited:Change Over Time
Of the six approaches described in this chapter, each has its own genealogy, disciplinary roots, instructions for
use, and challenges in application. It is perhaps not surprising that all six have evolved as new adherents have
adapted them to fit changing times and research interests. As mentioned earlier in this chapter, Glaser and
Strauss went their separate ways in their interpretations of grounded theory, and phenomenological analysis
took different forms dependent on the disciplinary proponent. Three major trends have occurred over time
that transformed qualitative methods in the 21st century.

First, and perhaps most significantly, several qualitative approaches were influenced by the rise of
constructivism and the postmodern debates that ensued after the 1980s. This influence, championed by
leading qualitative methodologists such as Norman Denzin and Yvonna Lincoln, manifested itself in
reflexivity and constructivism as well as postmodernist criticism. With regard to the former, narrative and
phenomenological approaches had little distance to travel with their predisposition toward social constructions
shaped by language or by the reflexive recounting of one’s experiences.

Ethnography underwent tremendous change wrought by criticism from native peoples, the disappearance of
traditional societies and postmodernist self-doubt. Thus, straightforward description of an assumed reality in a
faraway culture (with the investigator remaining invisible in the telling) gave rise to deeper interpretations and
multiple realities conducted closer to home. Along the way, ethnography evolved in new directions: on the
one hand introspective and on the other hand experimenting with new forms of representation and criticism.
Ethnography shifted “from participant observation to the observation of participation” (Tedlock, 2000, p.
465).

Multiple genres flowered, including critical ethnography (Kincheloe & McLaren, 2000), auto-ethnography (Ellis
& Bochner, 2000; Jones, Adams, & Ellis, 2013), performance ethnography (Denzin, 2003; McCall, 2000),
feminist ethnography (Tedlock, 2000), and institutional ethnography (Smith, 2005). Turning the first-person
account inward to the researcher’s own experiences, auto-ethnography brings out the richness of the research
experience and the interplay of emotions and positionality. Auto-ethnography and performance ethnography
introduced poetry and memoirs as well as literary writing (Clough, 1998; Denzin, 2003). The journal
Qualitative Inquiry features such works explicitly. Institutional ethnography (Campbell, 2004; Smith, 2005)
spotlights organizations and the procedures, processes, and discourses experienced by participants. Developed
by feminist sociologist Dorothy Smith, the method adopts a critical stance in keeping with Foucault’s focus on
power relations and inequality.

Annells (1996) and Mills, Bonner, and Francis (2006) find a strong constructivist thread running through
grounded theory (even though it is usually the winner in the category of “qualitative method most likely to be
postpositivist”). While Corbin and Strauss (2007) stayed fairly neutral epistemologically, constructivist GT
was developed by Charmaz (2006), and Clarke (2005) introduced postmodernist thinking into GT using
situational analysis. Charmaz distinguishes constructivist from objectivist grounded theory, noting that the
former relies on interpretive frames, and the latter focuses on explanation and prediction. Objectivism

83
presumes that data have meaning without reference to the context or researcher’s role—both Strauss and
Glaser accepted this presumption as well as many GT researchers who “discover” theory and emergent
concepts as having a “real” existence (Charmaz, 2014). Constructivist GT fully integrates the researcher into
theory making and interpretation. Study participants have their own interpretations, but these are part of a
larger enterprise in which the researcher practices reflexivity to ensure the “situatedness” of the knowledge that
is produced.

Action and participatory research have strong foundations in pragmatism as they relate to solving real-world
problems (Levin & Greenwood, 2001). However, constructivist iterations of CBPR have emerged. Eng and
colleagues, for example, cite constructivism as their research paradigm in working with rural African
American communities in North Carolina (Eng et al., 2005). In summary, constructivism has found powerful
allies among leading qualitative methodologists and has gained influence in many qualitative approaches.

And yet the constructivist and critical turns have not come to dominate qualitative research. As mentioned in
Chapter 1, postmodernist ideas generated pushback from some qualitative researchers (Atkinson, 2005). As
auto-ethnography and performance ethnography yield personal memoirs and poetic reflections, the scope
narrows and turns inward. For researchers interested in broader social, economic, and political concerns,
starting (and often ending) at the boundaries of personal experience misses a vital opportunity to engage.

A second major trend has been the blurring of boundaries across the six approaches and growing convergence
in some areas such as purposive sampling and thematic analysis. This phenomenon is not necessarily new, as
boundary maintenance within qualitative inquiry has never been a priority. Yet recent years have brought a
pragmatic willingness to mix and match methods (more on this in the next section of this chapter).

A leader in blending narrative research with other qualitative methods, Cheryl Mattingly (1998) examined
occupational therapists in hospital settings, their interactions with patients revealing the power of narratives in
cultural constructions of illness. Such “therapeutic narratives” helped to reframe the experience of disability,
grounding it in patients’ perspectives and the dialogic interactions with clinicians. In her more recent book
Paradox of Hope: Journeys Through a Clinical Borderland (2010), Mattingly proposes a “narrative
phenomenology of practice” grounded in the experiences of low-income African American families
confronting serious illness within the “border zones” of urban hospitals where race, culture, and biomedicine
coexist in uneasy tension. At the analytic stage, there has been a converging tendency toward thematic
analysis, not only of the codes and categories commonly found in GT but of narratives, case studies,
ethnographic data, and phenomenological interviews (more on this in Chapter 6).

The third trend has taken place on a more down-to-earth level as some qualitative approaches have been
altered to fit time-limited circumstances. Qualitative researchers in low-resource settings naturally look for
ways to truncate the methods without losing rigor. Regrettably, some turn to focus groups as the answer in the
belief that group interviews will efficiently yield the same type and amount of information as individual
interviews. Focus groups can be extremely useful but not in this role.

More considered attempts to retain the strengths of ethnography led to developing rapid ethnographic

84
assessment (REA). Perhaps not surprisingly, most of the pioneering work on REA was done by international
health organizations in collaboration with anthropologists. Anthropologists led the way in developing as a
means of conducting research in nutrition, sanitation, family planning, and HIV/AIDS (Beebe, 2002;
Manderson & Aaby, 1992; Scrimshaw, Carballo, Ramos, & Blair, 1991). Like its parent method, REA is
used in culturally specific situations and is not exclusively qualitative. Key informant interviews, for example,
could be combined with a survey measuring nutritional intake.

The success of REA is enhanced considerably when one or more of the investigators have prior knowledge of
the local culture as well as the requisite methodological skills. It would be difficult to imagine, for example,
trying to start a family planning program in eastern Kenya (or East Los Angeles) without knowing a great
deal about the governmental agencies and health officers involved as well as local religious beliefs, marital
practices, and views on women’s roles and rights.

Much less work has been done on adapting time-limited applications of the other qualitative approaches.
Indeed, qualitative inquiry continues to be defined by a timeline that is unpredictable and often extended.
Although time-sensitive techniques such as REA are needed for public health research and program
evaluation, their salience rests on the sturdy foundation of qualitative methods that are developed and carried
out in the pursuit of meaning, not deadlines.

85
Mixing and Matching Qualitative Approaches: Risks and Benefits
As mentioned in the previous section, mixing and matching qualitative approaches has become common.
However, a regrettable form of mixing comes from a lack of understanding or clarity in presenting the
methods. It is not uncommon, for example, to see a qualitative study that is presented as phenomenological
but uses grounded theory coding (Sandelowski & Barroso, 2003). At other times, the lack of clarity stems
from terminological confusion associated with the different methods and a lack of consensus on what these
methods entail. Thus, some case studies are hard to distinguish from ethnographies because they adopt an in-
depth holistic perspective, and grounded theory and phenomenological studies often resemble one another in
their reliance on thematic findings. Some narrative researchers use thematic analysis or produce typologies
similar to case studies. Qualitative studies often end up in similar places but arrive there via different routes.

Maintaining the integrity of a particular method does not preclude using it along with others. Annells (2006)
refers to this as “turning the prism” via methodological triangulation (p. 59). Matters get complicated when
considering (1) the different points at which mixing may occur—from interpretive paradigm to overall
approach to specific method of analysis and (2) the extent to which one is concerned about paradigm and
method congruence. A good example of this can be found in a study by Newman, Fox, Roth, and Mehta
(2004) in which they used a side-by-side paradigm approach to study school shootings in Kentucky and
Arkansas (predating the Columbine, Virginia Tech, and Sandy Hook tragedies). Newman and colleagues
used both positivist and interpretivist lenses, drawing on “factual” data from forensic analyses and court
records and also analyzing transcripts of interviews that provided conflicting (and conflicted) accounts by
students, school staff, and parents of the shooters as well as the victims.

Mixing approaches and techniques can bring a new synergy. Although less common (and much more likely to
be deployed at the analysis stage, hybrid mixing is described by Fereday and Muir-Cochrane (2006) in their
blending of the inductive procedures of Boyatzis (1998) with “template style” coding (Crabtree & Miller,
1999) to study nursing performance. Similarly, Wilson & Hutchison (1991) and Beck (1993) followed a side-
by-side use of phenomenology and grounded theory as complementary and mutually enriching perspectives.

Mancini (2005) sequentially mixed qualitative methods by carrying out a grounded theory study of persons
suffering from severe mental illness, then selecting two of the interviews for discourse analysis. In another
example of mixing in sequence, Teram, Schachter, and Stalker (2005) conducted grounded theory interviews
with female survivors of childhood sexual abuse enrolled in physical therapy, then shifted to “pragmatic action
research” to transform the analyses and findings via working groups of participants and physical therapists.
The groups’ joint production of a handbook for health professionals created a deeper, more sensitive set of
guidelines for working with abuse survivors.

A few related caveats are pertinent here. First, incongruities can bring complications during the mixing of
qualitative approaches. Phenomenological analyses of grounded theory interviews would likely suffer from the
lack of deep attunement to meaning and lived experience (Wimpenny & Gass, 2000). Second, mixing carries
the risk of “method slurring” (Baker, Wuest, & Stern, 1992) wherein one or both approaches loses its integrity

86
and capacity to make a contribution. This is obviously a greater problem for hybrid than for juxtaposing
formats. Finally, not all mixing is done for the purposes of triangulation (i.e., contrasting and/or
corroboration). As will be discussed in Chapter 8, triangulation is not as straightforward as it sounds.

87
Introducing Strategies for Rigor and Trustworthiness
One of the most vexing questions in qualitative research centers on defining what is “a good, valid, and/or
trustworthy qualitative study” (Sandelowski & Barroso, 2002, p. 2). Glaser and Strauss (1967) addressed this
question with a chapter titled “The Credibility of Grounded Theory” in their seminal work; Lincoln and
Guba (1985) provided their own discussions of quality, and a number of others have followed suit (Cohen &
Crabtree, 2008; Inui & Frankel, 1991; Morrow, 2005; Seale, 2002). Like their quantitative colleagues,
qualitative researchers seek respect and legitimacy, their efforts deemed worthy of recognition and wider
dissemination. Still, consensus has been elusive on how to achieve this coveted state. Meanwhile, critics of
qualitative methods are emboldened by this impasse. How, they ask, can one trust findings from studies where
standards are shifting and subject to diverse interpretations?

The volatile landscape of qualitative inquiry virtually guarantees that opinions about rigor will differ, including
whether the term itself is appropriate (some qualitative researchers prefer verisimilitude). Pivotal to discussions
about quality have been different ideas about the role of subjectivity, the stance of the researcher, and who has
the authority and legitimacy to judge good versus bad qualitative research. Once distance and objectivity cease
to be operating principles, the researcher’s subjectivity is acknowledged and, to varying degrees, managed
through reflexivity, or systematic self-awareness. Lincoln and Guba’s concept of trustworthiness (1985) comes
closest to capturing this phenomenon of quality and accountability in qualitative research. A trustworthy study
is one that is carried out fairly and ethically and whose findings represent as closely as possible the experiences
of the participants.

Although Chapter 8 is dedicated to this topic in greater and necessary detail, rigor is introduced here for two
reasons. First, qualitative researchers do not adhere to quantitative criteria; inappropriate invocations of
reliability and validity divert attention from more relevant criteria. Second, qualitative study designs need to
pay heed to rigor from the earliest stages. Recommended strategies for rigor constitute a menu of options. Not
all are appropriate for a given study, but each merits consideration at the design phase. In no particular order
or ranking, these are prolonged engagement, triangulation, peer debriefing and support, member checking, negative
case analysis and audit trail. The intrepid reader is invited to skip to Chapter 8 to read more about these
strategies.

88
Qualitative Methods in Program Evaluation and Implementation Research
Qualitative methods in evaluation have a long history in educational research (Cook & Reichart, 1979; Guba
& Lincoln, 1981, 1989; Patton, 2002; Scriven, 1967) extending to other professions concerned with program
effectiveness—business, public health, public administration, social work, and so on. Program evaluation is set
apart not by the methods used but by the goals it fulfills. Thus, randomized experiments are still the gold
standard, but the realities of programs operating under uncontrolled and often messy conditions necessitate
methodological compromises.

Relying solely on quantitative methods risks losing an understanding of what is happening below the surface
(where many a program succeeds or fails in ways unbeknownst to the investigator; Padgett, 2015). Any
number of hidden effects may occur. A program may be found successful but not for the reasons assumed. It
may appear to be a failure according to the selected outcome measures, but it might have been deemed a
success by different methods. Narrow conceptualizations of success (e.g., symptom reduction) may overlook
what clients value more (e.g., social support), and positive outcomes may be an artifact of biased sampling or
measurement error.

Given the basic distinction between process and outcome evaluation, qualitative methods are generally
associated with the former—that is, the hows and whys of the program and its inner workings. Qualitative
methods are also suitable for formative evaluation where the primary goal is improving the program prior to
full implementation. Certain facets of a program are difficult to capture and quantify—staff morale, executive
decision making, cultural misunderstandings, and client perceptions, among others. In-depth interviews and
on-site observation can add significant and unforeseen insights into how a program is operating in real time
and under fluid, changing conditions.

The praxis of qualitative research—with its emphasis on interpersonal relationships, rapport, and trust—is
conducive to program evaluations where staff and clients are understandably wary of researchers entering their
domain. Listening empathically and taking the time to fit into program activities smooths the way to
acceptance and greater cooperation. Another advantage of qualitative methods lies in their contribution to the
evaluation’s findings. Administrators and policymakers can get lost in (or bored by) a thicket of statistics, but
vignettes and direct quotes make a point that is more easily grasped and appreciated. Individual success
stories, or lessons learned from failure, are powerful ways to get the message across.

As described in Chapter 1, implementation science has arisen in recent years to address what happens when
programs and interventions move from controlled testing to real-world conditions in diverse settings
(Damschroder et al., 2009; Palinkas et al., 2011). Many evidence-based practices and programs fail during
implementation for a variety of reasons, some known but many unknown. An estimated 90% of public youth-
service systems, including mental health, education, juvenile justice, and child welfare, do not use evidence-
based practices (Hoagwood & Olin, 2002). To what extent is this situation a result of poor translation of such
practices from their evidence-demonstrating phase?

89
Such concerns point to the need for contextual methods sensitive enough to capture what is happening behind
the scenes, not just on the stage. Though more often part of mixed methods than stand-alone, qualitative
methods are critical to implementation research (Palinkas et al., 2011). It is difficult to imagine any type of
implementation study or program evaluation that could not benefit from the qualitative perspective.

90
Summary and Concluding Thoughts
This chapter introduced six primary approaches in qualitative inquiry—ethnography, grounded theory, case
studies, phenomenological analysis, narrative approaches, and action or participatory research. There are
commonalities—nonformulaic, iterative designs; in-depth immersion with participants as well as data; insider
perspectives; and pattern recognition as a route to analysis and possible theory development. There are also
important differences arising from disciplinary influences as well as epistemological preferences. Not always
willing to settle for just one, researchers often mix and match qualitative approaches to achieve the most
suitable combination for their needs. Novice researchers are urged to read specialized texts and articles using
these varied approaches to get a sense of how investigators make the most of what each has to offer.
Qualitative inquiry is steeped in choices and decisions—a qualitative study can be seen as a series of critical
junctures in which the decision trail is rarely, if ever, foreordained.

91
Exercises
1. Go to Google Scholar or use your college/university access to academic journals and locate examples of
studies representing each of the six types of qualitative methods presented in this chapter. What types of
journals carry these methods? Download and print an article and bring to class for discussion.
2. How would you describe the strengths and limitations of each of the six approaches presented in this
chapter?
3. Consider the many options possible in mixing among the six qualitative approaches. Discuss in class
which appear most (and least) suitable for mixing.

92
Additional Readings

93
Ethnography
Agar, M. H. (1980). The professional stranger: An informal introduction to ethnography. New York, NY:
Academic Press.

DeWalt, K. M., & DeWalt, B. R. (2001). Participant observation: A guide for fieldworkers. Walnut Creek, CA:
AltaMira Press.

Emerson, R. M., & Fretz, R. I. (2011). Writing ethnographic field notes (2nd ed.). Chicago, IL: University of
Chicago Press.

Fetterman, D. M. (2010). Ethnography: Step by step (3rd ed.). Thousand Oaks, CA: Sage.

Hammersley, M., & Atkinson, P. (1995). Ethnography: Principles in practice (2nd ed.). New York, NY:
Routledge.

LeCompte, M. D., & Schensul, J. J. (1999). Designing and conducting ethnographic research (Ethnographer’s
toolkit, Vol. 1). Walnut Creek, CA: AltaMira Press.

Lee, R. M. (2000). Unobtrusive measures in social research. Philadelphia, PA: Open University Press.

Lofland, J., & Snow, D. A. (2005). Analyzing social settings: A guide to qualitative observation and analysis.
Belmont, CA: Wadsworth.

Madison, D. S. (2005). Critical ethnography: Method, ethics and performance. Thousand Oaks, CA: Sage.

Smith, C. D., & Kornblum, W. (Eds.). (1996). In the field: Readings on the field research experience. Westport,
CT: Praeger.

Spradley, J. (1979). The ethnographic interview. New York, NY: Holt, Rinehart, & Winston.

94
Grounded Theory
Charmaz, C. (2014). Constructing grounded theory: A practical guide through qualitative analysis (2nd ed.).
Thousand Oaks, CA: Sage.

Clarke, A. E., & Friese, C. (2015). Situational analysis in practice: Mapping research with grounded theory.
Thousand Oaks, CA: Sage.

Corbin, J., & Strauss, A. L. (2007). Basics of qualitative research: Techniques and procedures for developing
grounded theory. Thousand Oaks, CA: Sage.

Dey, I. (1999). Grounding grounded theory. San Diego, CA: Academic Press.

Glaser, B. G. (1992). Basics of grounded theory. Mill Valley, CA: Sociology Press.

Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory. Chicago, IL: Aldine.

Oktay, J. S. (2012). Grounded theory. New York, NY: Oxford University Press.

95
Case Study Analysis
Gerring, J. (2007). Case study research: Principles and practices. London, UK: Cambridge University Press

Hancock, D. R., & Algozzine, B. (2006). Doing case study research: A practical guide for beginning researchers.
New York, NY: Teachers College Press.

Mills, A. J., Derepos, G., & Wiebe, E. (2010). Encyclopedia of case study research. Thousand Oaks, CA: Sage.

Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.

Stake, R. E. (2005). Multiple case study analysis. Thousand Oaks, CA: Sage.

Yin, R. K. (2013). Case study research: Design and methods (5th ed.). Thousand Oaks, CA: Sage.

Yin, R. K. (Ed.). (2004). The case study anthology. Thousand Oaks, CA: Sage.

96
Phenomenology
Colaizzi, P. F. (1978). Psychological research as the phenomenologist views it. In R. Valle & M. King (Eds.),
Existential-phenomenological alternatives for psychology (pp. 48–71). New York, NY: Oxford University Press.

Giorgi, A. (1985). Phenomenology and psychological research. Pittsburgh, PA: Duquesne University Press.

Groenewald, T. (2004). A phenomenological research design illustrated. International Journal of Qualitative

Methods, 3(1), Article 4.

Moustakas, C. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage.

Polkinghorne, D. E. (1989). Phenomenological research methods. In R. S. Valle & S. Halling (Eds.),

Existential-phenomenological alternatives for psychology (pp. 41–60). New York, NY: Plenum.

Vagle, M. D. (2014). Crafting phenomenological research. Walnut Creek, CA: Left Coast Press.

van Manen, M. (2014). Phenomenology of practice. Walnut Creek, CA: Left Coast Press.

97
Narrative Approaches
Andrews, M. (2014). Narrative imagination and everyday life. Oxford, UK: Oxford University Press.

Clandinin, D. J. (Ed.). (2006). Handbook of narrative inquiry: Mapping a methodology. Thousand Oaks, CA:
Sage.

Clandinin, D. J., & Connelly, F. M. (2004). Narrative inquiry: Experience and story in qualitative research. San
Francisco, CA: Jossey Bass.

Cortazzi, M. (1993). Narrative analysis. London, UK: Falmer Press.

Czarniawska, B. (2004). Narratives in social science research. London, UK: Sage.

Elliott, J. (2005). Using narrative in social research. Thousand Oaks, CA: Sage.

Frank, A. W. (2010). Letting stories breathe. Chicago, IL: University of Chicago Press.

Gee, J. P. (2005). An introduction to discourse analysis: Theory and method. London, UK: Routledge.

Josselson, R., & Lieblich, A. (Eds.). (1995). Interpreting experience: The narrative study of lives (Vol. 3).
Thousand Oaks, CA: Sage.

Mattingly, C. (1998). Healing dramas and clinical plots: The narrative structure of experience. Cambridge, UK:
Cambridge University Press.

Riessman, C. K. (2008). Narrative methods for the human sciences. Thousand Oaks, CA: Sage.

ten Have, P. (2007). Doing conversation analysis (2nd ed.). London, UK: Sage.

98
Action and Community-Based Participatory Research
Cornwall, A., & Jewkes, R. (1995). What is participatory research? Social Science & Medicine, 41(12), 1667–
1676.

Israel, B. A., Eng, E., Schulz, A. J., & Parker. E. A. (Eds.). (2012). Methods in community-based participatory
research for health (2nd ed.). San Francisco, CA: Jossey-Bass.

Jones, L., & Wells, K. (2007). Strategies for academic and clinician engagement in community-based
partnered research. Journal of the American Medical Association, 297(4), 407–410.

Minkler, M., & Wallerstein, N. (2008). Community-based participatory research for health (2nd ed.). San
Francisco, CA: Jossey-Bass.

Reason, P., & Bradbury-Huang, H. (2013). Handbook of action research: Participative inquiry and practice.
Thousand Oaks, CA: Sage.

Stringer, E. T. (2013). Action research: A handbook for practitioners (4rd ed.). Thousand Oaks, CA: Sage.

99
Websites Related to CBPR
https://mailman.u.washington.edu/mailman/listinfo/cbpr (news and updates on funding for CPBR)
https://ccph.memberclicks.net/participatory-research (resources for CBPR)
http://depts.washington.edu/ccph/commbas.html (all-purpose resource site for CBPR)
https://mailman13.u.washington.edu/mailman/listinfo/cbpr (listserv for those interested in CBPR)
http://www.cbprcurriculum.info (information on CBPR curricula)
https://mailman13.u.washington.edu/mailman/listinfo/ccph-ethics (CBPR ethics discussion group)
https://mailman1.u.washington.edu/mailman/listinfo/communitypartnerlistserv (community partner
discussion group)

100
3 Designing the Study and Getting Started

101
Prologue: Addressing Ethical Considerations Early On
Chapter 4 will address ethics and human subjects protections in greater detail, but it is important to introduce
a few considerations during the getting started phase. Although qualitative studies seldom entail a serious risk
of harm for study participants, the sensitivity of their subject matter and closeness of relationships pose
challenges that should not be taken lightly.

The flexibility of qualitative inquiry means that unforeseen ethical dilemmas can arise at any time, but some
issues can be anticipated and dealt with in advance. These include avoiding coercion, maintaining
confidentiality, and tending to sensitivity and emotional issues. One’s mode of recruitment and data collection
can be coercive if reliant upon powerful gatekeepers or inconsiderate of participants’ inviolate right to refuse.
Moreover, getting informed consent does not preclude the potential for breaching confidentiality, unintended
though it may be. Qualitative study participants reveal a great deal about themselves, and the researcher must
remain vigilant in ensuring that such revelations are kept in strictest confidence during data collection and
later on when reporting the results. Relatedly, the intimacy and intensity of qualitative inquiry can leave
participants emotionally vulnerable. For some studies, it may be advisable to build into the research design a
mechanism for referrals to psychological counseling in the unlikely event that intense emotions do not resolve.

102
Flexible and Iterative Designs
The idea of having a design in qualitative research is relatively new (and undoubtedly anathema to those who
eschew procedural specification as dampening creativity). Traditional ethnographers and grounded theorists
simply studied what interested them and let the contours of the study emerge over time (lots of time). There
are a few reasons why this is not common today. First, truly open-ended inquiry, with no prior
methodological grounding, is an extreme rarity due to a lack of time and resources. Some early anthropologists
were funded by (or embedded within) colonial governments; others had family wealth sufficient to sustain
them for years in the field. Relatedly, human subjects protections and oversight by thesis and dissertation
sponsors have raised the ante in terms of description and transparency of a study’s methods. Research careers
are built on trustworthiness that is demonstrated in advance as well as earned.

Quantitative designs are analogous to architecture or engineering; blueprints are carefully drawn in advance
down to the smallest specification and then followed closely. Qualitative research designs need to be
systematic, transparent, and as rigorous as possible while being flexible and iterative. Miles and Huberman
(1994) refer to the “second chance” (p. 38) afforded by the flexibility and groundedness of qualitative
methods. Transparency entails keeping meticulous records of what is done and thus maintaining a degree of
accountability for decisions made along the way.

Some researchers begin with broad study questions that spring from intellectual curiosity, prior theoretical
frameworks, personal experience, and/or a commitment to social change. Study questions are distinct from
research questions in being more broadly conceptualized and not directly researchable as stated. An example of a
study question might be, How do refugee families adjust to postwar resettlement? Research questions might
include, How are family relationships affected by displacement? How are decisions made about returning or
staying in the new location? Are there signs of resilience and strength that distinguish some families from
others?

103
Choosing the Topic and Making the Argument
Scholarly work is built around making an argument and using rhetorical devices and available evidence to
support it. Researchers must anticipate the “so what?” question because their interest in a topic may not
resonate beyond their immediate circle of friends and family (if at all). The more personal the interest, the
greater the risk of sounding and being self-absorbed. On the other hand, the topic may be unfamiliar to the
researcher but written about extensively by others. Either way, the “so what?” question compels an explanation
for why the new study is needed.

Researchers interested in exploring unknown (or relatively unknown) terrain have a distinct advantage
(anthropology made its reputation in this way). However, an esoteric topic can erode this advantage. Most
research topics have been previously studied and it is incumbent on the researcher to raise new questions that
address a gap in knowledge needing to be filled. For researchers seeking external funding, the stakes are the
highest, that is, one’s powers of persuasion and rhetorical reach are fully put to the test.

104
Reviewing the Literature
Although more often assumed than articulated, the fund of knowledge brought to bear by the researcher can
make the difference between an innovative study and a ho-hum effort. Building upon previous work sets the
stage for deeper description, conceptual development, and refinement of ideas. A literature review is both
descriptive and evaluative. The writer should not shy away from substantive controversies or debates but also
should not betray a strong bias. Ultimately, what matters is the weight of the argument and its foundation in
empirical research. Most researchers start out with a funnel approach, that is, casting a wide net using
database and key word searches then gradually narrowing the scope to be relevant and manageable.

Novice researchers (and quite a few experienced ones) find the literature review one of the most arduous tasks
in a study. How far does one go in casting the net? Internet searches can yield literally thousands of articles,
book chapters, and abstracts, not to mention the expansive “gray literature” of unpublished reports. Once the
net is cast, how does one decide if a study is relevant or tangential? Then, how does one thread an argument
into the narrative and make sure it leads to (or funnels the reader toward) the study’s research questions?

It helps to develop an outline with headings that structure the review. Typically, a description of the nature
and extent of the problem comes first to establish a rationale for doing the study. Topics or problems having
urgency and wide impact are easier to rationalize than mundane or idiosyncratic topics. If appropriate, the
next section of the review covers relevant theories and frameworks, critically appraising them and mining
them for concepts and schemas that help frame the topic of study. Even though theories are treated gingerly
in qualitative research, it would be negligent to overlook their contributions. The next section goes to the
heart of the matter—a critical review of empirical research that highlights gaps in knowledge and puts forward
the need for the proposed study. Some literature reviews have separate sections for quantitative and qualitative
research, but this is not mandatory.

The length and organization of a literature review depend on the study’s purpose. Dissertations have few if
any restrictions on length and high expectations regarding comprehensiveness and theoretical import. In
contrast, evaluation reports present background information in summary form. What distinguishes a
qualitative literature review from its quantitative counterpart is a lack of conceptual closure. Conceptual and
theoretical doors are left open wide enough to permit new ideas and serendipitous findings to emerge.

105
Developing a Conceptual Framework
Literature reviews contain (and are often organized around) key concepts that are touchstones for the study
and its research questions. A qualitative study’s conceptual framework is not a contractual obligation. Rather,
it is a guiding influence that ensures the study will transcend description (no matter how rich and compelling).
As discussed in Chapter 1, theories and concepts play an integral but unpredictable role in qualitative inquiry.
The various qualitative approaches treat a priori concepts differently, but all assume some reference to the
world of ideas swirling around a particular topic. Conceptual frameworks are invoked in the literature review,
lightly applied during analysis, and revisited during the interpretation of the findings.

106
Formulating Research Questions
Qualitative research questions pave the way for a qualitative study. This may sound trite, but the sentence
contains two operative words. Study means that the questions must be researchable, not broad abstractions.
Qualitative means that the study must be inquisitive and open minded. We humans think about things in
complex nonlinear ways, and the most interesting ideas rarely lead to researchable questions without a lot of
conceptualizing, word crafting, and feasibility balancing. This journey begins with intellectual curiosity
(sometimes aided by personal experience). It proceeds toward further expansion, then refinement and
contraction during the literature review, the latter ending with a few finely honed questions that can only be
answered by using qualitative methods.

Of course, phrasing and intent will vary with the study’s approach. Take, for example, an interest in studying
young women who are HIV positive and engaged in sex work. The reason for doing so (for a researcher in
social work or public health) is fairly obvious. What is less obvious is how to go about such a study. The
following are some examples of research questions matched to the type of method.

Ethnography: Are there tacit values, beliefs, and practices that characterize a local “culture” of sex
workers? If so, how do these affect the women’s decisions about personal health and HIV prevention?
How do the women negotiate relationships with their clients and with fellow sex workers? What is daily
life like for them?
Grounded Theory: How do women with HIV balance sex work with other life demands? Are there
common elements to their experiences that can be identified as part of a grounded theory of having HIV
among sex workers?
Case Study: What life events (childhood or recent) led the women to sex work and to becoming HIV
positive? How are their life stories similar or different to that of other women of their age?
Phenomenology: What is the lived experience of being HIV positive? What is the lived experience of
sex work? What are the essential elements of the life worlds of these women?
Narrative Analysis: What stories are embedded in their narratives? How do these women “voice”
themselves and others in their social networks? What do these narratives reveal about exposure to HIV
and AIDS prevention in sex work?
Action or Participatory Research: What are the needs of these women as they perceive them? How can
researchers join with them in a partnership to conduct research that addresses these needs?

Any of these may incorporate theoretical and conceptual perspectives. For example, feminist researchers might
include a focus on the constraints of gender roles in sex work and of sexism in AIDS treatment and
prevention. Case studies might adopt a life course perspective for each woman and examine childhood
adversity as well as the impact of historic events such as new antiretroviral medications.

The pitfalls to crafting qualitative research questions include a tendency to think and write quantitatively. It
helps to project into the future what the study would look like based upon the research question as posed:
Does answering the question lead to insider perspectives and rich information? Research questions phrased as

107
cause-effect or correlational statements are not a good fit, for example, “what is the relationship between
partner violence and alcohol abuse?” or “what is the impact of court-mandated drug testing on recidivism?”
Qualitative studies are not meant to answer cause-effect questions even though their findings often help
researchers understand the “how” or “why” of causal mechanisms.

The discerning reader will notice here and elsewhere an avoidance of hypotheses. Some qualitative
methodologists (Miles & Huberman, 1994; Tashakkori & Teddlie, 2010) assert that hypotheses can be tested
and confirmatory or explanatory analyses conducted using qualitative data. Grounded theory can be described
as an inductive–deductive feedback loop in which hypotheses or hunches are tested as the conceptual model is
built (although the study begins with questions, not hypotheses). While not ready to throw out hypothesis
generation and testing during analysis, expending too much effort in that direction undermines a unique
strength of qualitative methods.

108
Designing the Study
The word design sounds almost too orderly for the iterative process that unfolds in most qualitative studies.
Whereas quantitative researchers share a common taxonomy of designs (experimental, quasi-experimental,
time-series, etc.), no such uniformity exists in qualitative inquiry (Maxwell, 2012). Most qualitative
researchers opt for a straightforward description of what they plan and how they plan to do it, using as many
descriptors as are applicable (case study, ethnography, phenomenology, etc.).

Qualitative designs are distinguished by their recursiveness and flexibility, often weaving back and forth
between research questions, data collection, and data analysis. In this fashion, the researcher may reformulate
research questions based on new findings, seek new samples of respondents, or pose new questions to existing
study participants. Similarly, data analyses can precipitate the collection of additional data.

In quantitative designs, a formulaic approach puts the emphasis on minimizing external “noise” and threats to
validity (i.e., “What you see is what you get”). In contrast, qualitative researchers pride themselves on viewing
noise as an inevitable and even welcome part of naturalistic studies. Yet they must also convince their audience
that they can (and will) produce credible, trustworthy findings. This means offering an explicit message of
“here is what I/we plan to do” tempered by the conditional message “these plans may be modified, but all
analytic decisions will be justified and made transparent.” Of course, like all complex endeavors, the devil is in
the details.

109
Questions Posed (and Needing Answers) by Qualitative Research Designs
Box 3.1 shows a series of questions typically posed by qualitative designs. Questions with relatively
straightforward answers have them provided in parentheses. Questions without immediate answers are
asterisked as “it depends.” Perhaps not surprisingly, these questions cannot be answered without reference to
the specific method being used as well as the scope of the study. Decisions about sampling, data collection,
and data analyses are usually method specific, for example. Regardless, a qualitative research proposal should
aim for as much specificity as possible. To be sure, overattention to detail may drain away the creativity that
lends qualitative findings their strength and longevity. But creativity need not preclude planning ahead.

110
 Box 3.1 Questions Posed by Qualitative Research Designs
“How many?” questions

How many study participants are needed?*

How many interviews per participant? (Whenever possible, there should be at least two.)
How many study sites are needed?*
How many weeks, months, or years are needed to complete the study?*

*It depends.

“Should I?” questions

Pilot test the interview guide? (Yes!)

Pay participant incentives? (Yes, if you can afford them.)
Have comparison groups? (Only if your topic requires it.)
Collect observational data if it is an interview-based study? (Yes, if possible.)
Identify the strategies for rigor to be used? (Yes, but not all are needed or appropriate.)

“When?” questions

When should I start analyzing the data? (As early as possible.)

When is there enough data to stop collection? (Usually at “saturation” point.)
When should mentoring and supervision be sought? (Early and often.)

“How?” questions

How do I address ethical concerns? (More on this in Chapter 4.)

How do I sample study participants? (This is discussed later in this chapter.)
How do I leave the field and end data collection? (See Chapter 6.)
How do I analyze and interpret the data? (See Chapters 6 and 7.)
How do I write and present the findings? (See Chapter 9.)

Baker and Edwards (2012) polled 14 qualitative methods experts and five early career researchers, asked them
the “how many interviews?” question, and compiled the responses. The answers “until saturation” and “it
depends” were common with the acknowledgement that saturation is optimal but unrealistic as an answer to
thesis advisors and research funders. The “it depends” answer is a nod to variations in time and resources with
fewer interviews expected when these are in short supply.

Interestingly, the Baker and Edwards study focused on sample size rather than number of interviews per
participant. Most of the experts consulted were uncomfortable with any set responses regarding sample size
(one jokingly stating that 20 for a master’s thesis and 50 for a dissertation is about right). Other experts
advised the researcher to reflect on the study’s epistemology as well as its audience (academic or general) and
those who will sit in judgment of the study (funders, thesis advisors, peers; Baker & Edwards, 2012).

111
The Element of Time
Flick (2004) notes the importance of time in design decisions. If concerned with temporal change, qualitative
studies may be retrospective (e.g., life histories) or prospective, using longitudinal designs. When change is not
the focus, qualitative studies may be “snapshots” (Flick, 2004, p. 148) or cross-sectional in design.
Anthropologists, for example, talk about the “ethnographic present” as the compression of the lengthy period
(usually a year or more) when they conducted their research in the field.

While cross-sectional designs may entail data collection conducted over several months (or even years),
longitudinal designs are focused on processes of change (Saldana, 2003). What typically makes a study
longitudinal is (1) its reliance on two or more waves of interviewing separated by a specified time interval and
(2) its ulterior motive of examining change over time. Iversen and Armstrong (2006), for example, conducted
ethnographic interviews with low-income families over a 5-year period to portray their ongoing struggle with
declining economic fortunes.

112
Specific Aspects of Qualitative Designs
Notwithstanding the looping back and forth that characterizes most qualitative designs, the description itself
is fairly linear. Building on the questions posed in Box 3.1, the following are primary items to consider:

1. Which qualitative method(s) will be used? Include a rationale for this choice.
2. If the study is longitudinal, describe this, including procedures for retaining participants.
3. How many participants will there be, and how will they be sampled?
4. List the inclusion/exclusion criteria for eligibility.
5. How, where, and by whom will participants be recruited?
6. How will informed consent be obtained and human subjects protections maintained over the course of
the study?
7. What types of data collection will be pursued?
8. How many interviews will be conducted per participant?
9. About how long will interviews last, and where will they take place?
10. How (if at all) will incentives be paid, and how much will they be?
11. How will data be managed and transformed, including transcription?
12. How will data be analyzed (including kind of software to be used)?
13. Which strategies for rigor will be used?
14. How will findings be presented and disseminated?
15. Give a timeline for completion of all study tasks.

Creating a timeline of tasks or a schema is an excellent way of visually displaying what needs to be done and
when, even if presented with a caveat regarding the flexibility that may accompany sampling or data
collection. All descriptions should liberally cite the relevant experts in the literature and offer thorough
descriptions and rationales.

One modest but essential way to enhance a qualitative study is to build in a pilot study. Pilot studies help
smooth out wrinkles in the study’s execution and refine its protocols. Little has been written about this in
qualitative texts, in large part because of the unpredictable nature and absence of fixed protocols. A pilot study
in no way foretells all of the complications that lie ahead, but it often results in improved questions for the
interviews or better procedures for data collection.

113
Box 3.2 Lost in Translation? Design Considerations for Non-English
Language Studies
In the early days of ethnographic fieldwork, the anthropologist was expected to become fluent in the local dialect or, at the least, hire
a local translator. As qualitative research entered a boom period in the 1980s, its adoption by the professions (educators, nurses,
social workers) meant that most studies were conducted closer to home and language barriers were not an issue, that is, research was
done by English speakers with English speakers. Recently, cross-language studies have reemerged as qualitative researchers have
become increasingly interested in diverse peoples and places. Confronting language differences early on means building into the
study mechanisms to maximize accuracy and fidelity in translation. There are a few guidelines for doing this, but common sense
would dictate that, at a minimum, all relevant materials—consent forms, recruitment flyers, and interviews guides—are carefully
translated (and back-translated) to ensure accuracy and consistency.

The stickier wickets come with data collection and analysis and the cultural context in which the non-English language is embedded.
Some researchers hire and carefully train someone to conduct the interviews in their native language—this is preferable to the stop-
and-start disruptions of translation during an interview. A researcher who is both linguistically and culturally fluent might conduct
the analyses in the participants’ language and use English only for the final write-up and publication. More commonly, the
researcher uses a fluent bilingual speaker to listen to the audio recordings and translate the interviews into English for subsequent
analysis. Words or idiomatic phrases not easily translated into English are kept intact (usually in italics) and their meaning explained
in brackets or footnotes. This is obviously a compromise—and some might argue that qualitative studies dependent upon deep
understanding, for example, phenomenological or narrative analyses, are not suited to cross-language translation.

114
Using Multiple Qualitative Methods at Different Levels
As discussed in Chapter 2, the topic of interest and scope of the study may lead to using more than one
qualitative method and technique. It is helpful to distinguish between mixing approaches (grounded theory,
ethnography, etc.), methods of data collection (focus groups, in-depth interviews, observation, etc.), and
analytic methods (coding, case study analysis, discourse analysis, etc.). Comingling at the methodological
approach level implies a study with sufficient expertise and coordination to manage the different tasks.

Within a particular qualitative approach, it is not uncommon to use multiple types of data collection.
Ethnography supplements its trademark participant observation with key informant interviews. A grounded
theory study might use interviews from focus groups as well as individuals. A life history might go beyond
first-person accounts to draw on participants’ documents such as diaries, photographs, newspaper clippings,
and so on. Qualitative approaches such as narrative and phenomenological analyses do not lend themselves as
easily to multiple data formats given their almost exclusive reliance on interviews and narrative. Mixing may
occur sequentially as well as concurrently. In any event, these decisions should be reflected in the study’s
design and its description.

115
Sampling and Selection Strategies
Qualitative researchers sample places and events as well as people. Sampling may be multilevel, beginning
with larger units such as schools or agencies and then proceeding to selected individuals at these locations
(e.g., teachers or staff). Qualitative studies tend to rely heavily on people who are articulate and introspective
enough to provide rich descriptions of their experiences. Interviews that produce sketchy answers from
disinterested respondents are poor sustenance for a study; it is not unusual to discard an occasional interview
as unusable. And yet, with the possible exception of young children and severely demented persons, few, if
any, individuals should be ruled out as potential study participants. Thus, stereotyped assumptions that
persons with severe mental illness are incapable of participating in qualitative interviews have been proven
unfounded time and again.

Of course, the ultimate decision about sampling should be driven by the study’s research questions and goals.
As a general rule, qualitative researchers use purposive sampling—a deliberate process of selecting respondents
based on their ability to provide the needed information. As Miles and Huberman note (1994), qualitative
sampling is done for conceptual and theoretical reasons, not to represent a larger universe.

Purposive sampling implies that a researcher interested in how cancer patients cope with pain will seek out
respondents who have pain rather than randomly sample from an oncologist’s patient roster. As such, it
should not be confused with convenience sampling, or selecting respondents based solely on their availability.
Commonly used in clinical research (where ready access to specific types of patients overshadows concerns
about nonrepresentativeness), convenience sampling is generally antithetical to the aims of qualitative
methods. Convenience may lead a researcher to a particular site (e.g., a domestic violence shelter where he or
she has volunteered in the past), but this should be done only if that site is most appropriate for the study.
Even when an appropriate site is available, the method for recruiting and selecting study participants should
be purposive, not one of convenience.

Patton (2002) describes various types of purposive sampling. Some of these are suited to the outset of a study,
and others are for later use. Initial sampling techniques are described here:

Extreme or deviant case sampling looks for cases that illuminate the “outer edges” of a phenomenon, for
example, persons with major depression who have had multiple suicide attempts.
Intensity sampling is similar to the above but the cases are not as unusual—persons with severe
depression, for example.
Maximum variation sampling captures heterogeneity across the sample population (e.g., recruiting cancer
survivors who had all types of adjuvant therapy [chemotherapy, radiation, etc.] as well as those who
rejected such treatments in favor of alternative remedies).
Homogeneous sampling is the opposite of the previous point, for example, narrowing the sample to
include only those cancer survivors who rejected adjuvant therapies.
Typical case sampling recruits average members of the population (e.g., parents of autistic children who
typify a profile of this population).

116
 Critical case sampling involves recruiting to illuminate the extremeness of a situation, for example,
parents whose abuse resulted in the death of their child.
Criterion sampling is selecting cases that exceed some criterion or norm (e.g., new mothers who score
above a clinically significant level on a postpartum depression scale). A variant of this, nominations
sampling, asks knowledgeable persons to name or select eligible persons based on the study criteria.
Snowball sampling is used with isolated or hidden populations whose members are not likely to be found
and cooperate without referral from others in their network. Examples include gang members, IV drug
users, or members of a religious sect. A quantitative variant of snowball sampling, respondent-driven
sampling (RDS), was developed to assist researchers in gaining access to hard-to-reach populations to
study AIDS transmission, IV drug use and other hidden behaviors (Heckathorn, 1997). Using chain-
referrals and “steering incentives” for recruitment, RDS has been shown to yield larger and more
representative samples despite having origins in snowball or convenience samples (Salganik &
Heckathorn, 2004).

Analysis-driven sampling techniques occurring later on in a study may include any of the initial sampling
techniques previously mentioned, along with the following:

Theoretical sampling, which occurs when inductively derived analytic concepts are used to guide the
choice of additional participants. For example, a grounded theory study of young adults leaving foster
care finds that those who reconnected with their birth family had more positive views of their future.
Further sampling of this population would yield a deeper understanding of why this might be the case.
Confirming or disconfirming sampling, which takes the logic of theoretical sampling a step further to seek
out cases to test the validity of the grounded theory. Building on the above example, additional sampling
of young adults who have not reunited with their birth family might identify whether such an association
is born out and worth pursuing. A variant of negative case analysis, this type of sampling opens up
preliminary findings to potential disconfirmation.

These categories offer the researcher useful terminology on specific sampling techniques. Not surprisingly,
random sampling is a rarely used option in qualitative research. Aside from the fact that many study
populations do not have a sampling frame from which to randomly select, the need for small but meaningful
samples makes random selection techniques less appropriate.

Sampling strategies should, whenever possible, identify inclusion and exclusion criteria to set boundaries on
who is and is not eligible. Researchers must make special provisions if they seek to interview anyone who is
under age 18 or a member of a vulnerable population (pregnant women, severely mentally disabled persons,
and prisoners). Sharing the study’s inclusion/exclusion criteria is essential for those helping to recruit for the
study, for example, staff at an agency or clinic who are asked to pass out flyers or otherwise help to identify
eligible participants.

Meeting the study’s sampling goals requires tact, persistence, and foresight. Two related observations are
pertinent here. First, sampling strategies may change in response to study needs. One might, for example,
start out with maximum variation sampling and then turn to a more targeted technique such as deviant case

117
sampling. Second, the researcher should take steps early on to ensure that flexibility in sampling can be
pursued and sampling goals attained. If recruitment is being carried out among certain types of clients or
patients, it is advisable to obtain assurances (and evidence) of sufficient accrual from their program
gatekeepers or else expand recruitment to more sites. Experienced researchers often have agonizing accounts
of how an intake coordinator promised a veritable flood of eligible study participants only to be confronted
with a trickle when the study began.

118
Sample Size Considerations
Different qualitative approaches have differing sample size considerations, and there are no hard and fast
rules. Case study analyses tend to have small samples—a single case may suffice in some instances. Similarly,
phenomenological studies aim for depth—sample sizes of 6 to 10 participants are common, but the numbers
may be somewhat larger if resources permit. Grounded theory studies tend to have larger sample sizes,
although still usually well short of a quantitative sample.

A few suggestions are helpful to remember: (1) the smaller the sample size, the more intense and deep are the
data being collected; (2) larger sample sizes are needed for heterogeneity, smaller sizes for homogeneity; and
(3) avoid sacrificing depth (length of interviews or number of interviews) for breadth (number of participants).
If, for some reason, you are unable to conduct more than one interview with each study participant, a larger
sample may be desirable: (4) larger numbers need not be shunned as long as the study has sufficient resources
and follows suggestion #3.

An assertion that saturation alone will be used to determine sample size is not likely to be taken seriously.
Various overseers—thesis advisors, human subjects committees, and potential funders—expect specificity, so
it is important to give a number even if the sample size might grow or shrink later on. Continuing data
collection beyond saturation is not a punishable offense, and it is better to expend the extra effort than risk
underreach. Whatever the proposed sample size, there is no need to apologize for it as long as it fits the
study’s goals. Small samples are not a limitation unless they portend thin data.

119
Recruiting and Retaining Study Participants
Qualitative researchers go where respondents are rather than the other way around. These locations could be
clinics, beauty salons, day care centers, churches, homeless encampments, or any site where would-be study
participants can be found. Gaining the cooperation of intermediaries such as gatekeepers is often critical.

Researchers in the helping professions are naturally drawn to studying problems of the populations they serve.
However, this can be risky if they recruit study participants solely through service settings. Such reliance
excludes nonservice users and thus skews the sample toward the more severe or chronic cases. Of course, this
may be the only way to access the population (or it may be the proper site given the study’s goals), but intrepid
researchers pursue alternative routes to recruitment such as advertising or snowball sampling. Consider a
qualitative study of veterans who suffer from posttraumatic stress disorder (PTSD). As a social worker, you
have access to a treatment program for these men and women. This could be your only opportunity to gain
such access, but it comes at a price since veterans with untreated PTSD are inherently of interest to
researchers and practitioners.

There are a variety of strategies for recruiting respondents. Human subjects committees are especially fond of
advertising for volunteers to directly contact the researcher because it is entirely volitional. Cooperating sites
and gatekeepers can help in distributing flyers or mailing introductory letters to potential participants. The
researcher may make guest appearances at regular group meetings to describe the study and invite
participation. At all times, he or she should have a script ready, either written or verbal, giving a brief
description of what the study is about and what participation will entail. Attractive flyers with these points in
bulleted form are the most efficient way to get the message out.

Casting the net too widely can drain time and resources, as the researcher must screen candidates for eligibility
according to the study criteria. Rejecting many would-be participants also causes disappointment and bad
feelings. Obtaining the optimal sample size is more successful if the researcher starts local then adjusts the
sampling strategy as the need arises.

Retaining study participants is a key element of study success. While essential for longitudinal studies, study
retention pertains to all qualitative research because multiple interviews with each participant are optimal, and
these may be extended over a period of weeks or months. The relational aspects of qualitative research (along
with smaller sample sizes) confer an advantage in this regard, but there are specific techniques that can help.
These include asking for contact information and paying incentives as well as giving participants a telephone
number or e-mail address where they may reach out with questions or concerns.

120
How Many Interviews? Issues of Quantity and Quality
There are qualitative studies in which one interview per participant is the only option and even a few where
this is preferable. As a general rule, however, it is optimal to conduct at least two interviews per person and
allow for more than two when possible (Seidman, 2006). A single interview starts the process and builds
rapport and forward momentum for the next interview(s). Multiple interviews contribute to evidentiary
adequacy (Erickson, 1986; Morrow & Smith, 1995).

The rationale for a single interview should be ethical rather than the researcher’s convenience. Patients in
hospice care or other vulnerable groups might find more than one encounter to be burdensome or intrusive
(though this should not be assumed automatically). Follow-up interviews fill in missing information, but they
are also important as a venue for pursuing leads from earlier interviews. In grounded theory studies, this may
occur under the rubric of theoretical sampling where reinterviewing or sampling new participants is an
anticipated stage in theory development. Repeated interviewing brings engagement that is vital to a qualitative
study. When using single interviews, the onus is on the researcher to make the most of these encounters.

121
Getting Permission(s) and Obtaining Rapport
After spending a considerable amount of time conceptualizing and designing the qualitative study, the
cerebral gives way to the social. For most researchers, this is energizing, the time when the planning finally
gets put to the test. Qualitative studies may be multisite, single-site, person-centered, or all of the above. It is
usually necessary to seek permission from various gatekeepers in order to start the research and recruit
participants. In the NYSS and the NYRS, we recruited participants from homeless service programs and
needed formal permission from their executive directors. The clients we recruited were themselves gatekeepers
whose consent was required in order to interview their case managers at the agencies as well as significant
others that they nominated for ancillary interviews.

Gaining entry is usually not an issue if one recruits directly through a published ad or via the Internet. But it
can become complicated when there are multiple sites and levels of authority needing attention (including
human subjects committees). Protracted negotiations are sometimes needed to assuage concerns, their success
dependent upon a willingness to accept researchers poking around. Some bureaucratic entities are suspicious
by nature (especially if subject to negative media reports), some are wary but workable, and others are
welcoming and enthusiastic.

It is best not to take anything for granted. Study participants and communities have legitimate concerns about
the potential for human subjects abuse or negative portrayals, giving rise to a “nothing about us without us”
mentality (Malone, Yerger, McGruder, & Froelicher, 2006). Qualitative researchers have to accept the fact
that the word research may conjure images of human guinea-pig experiments and exploitation. Substitution of
terms such as research project or study participant instead of subject is less distancing and more consonant with
the realities of a qualitative study.

The gatekeeper could be an agency administrator, corrections officer, school principal, or any other authority
figure whose support is needed to gain access. In approaching such individuals, the researcher should be direct
and forthright about the study’s overall goals and the human subjects protections being offered. The study’s
benefits should not be overstated; it is also important to inform gatekeepers and others about the time-
consuming and involved nature of qualitative research. If an incentive or other reimbursement is being offered,
it should be identified as such and not used as leverage or coercion. One advantage of qualitative research in
these situations is its user-friendly, low-tech approach. Still, it is advisable to get a head start in learning about
the settings and their gatekeepers, even those whose cooperation involves little more than posting a
recruitment flyer on a bulletin board.

Announcing the study refers to laying the groundwork for observation and/or interviewing specific to a site.
As with the usual procedures of informed consent (discussed in greater detail in Chapter 4), this involves full
disclosure of one’s identity, the purpose(s) of the study, its voluntary nature, and the protection afforded by
strict confidentiality. If field notes will be taken, this must also be noted so that everyone knows that they
might be observed. Other expectations of cooperation, such as help in recruiting, participation by staff in focus
groups, and so on, must be clearly specified. For a clinic or agency, this can usually be handled by attending

122
staff meetings to explain the study and answer questions. For larger entities, such as multisite programs or
school districts, letters of approval by gatekeepers can be sent to local authorities to smooth the way. Studies
of neighborhoods often depend on cultivating the goodwill of civic leaders, community boards, and local
authority figures.

Observational studies bring special considerations. With the exception of busy and open public spaces, the
researcher needs to broadcast information about the study as widely as possible. During intensive participant
observation, the researcher’s prolonged presence gives ample opportunities for such self-identification. For
sporadic or less intense involvement, having flyers and gatekeeper permissions handy and being available for
questions go a long way toward allaying suspicions. Of course, any contacts that lead to formal interviewing
will trigger the need for informed consent.

If a site is being asked to provide assistance with multiple tasks over a period of time, a memorandum of
understanding (MOU) is helpful at the outset to specify responsibilities of the respective parties. Although not
as necessary in qualitative research as in clinical trials (where randomization, wait list controls, and fidelity
measures often intrude on program operations), an MOU can be a reference point for everyone to clarify roles
and reduce misunderstandings. A simple one-page, bulleted list of responsibilities for the site and the
researcher is a good start. Of course, an MOU should not lock the research in too stringently because
protocols need to remain flexible.

Once all formal permissions have been obtained, what James Spradley (1979) refers to as the “rapport process”
(p. 78) takes center stage and continues until the study is over. Rapport can begin on a number of levels,
depending on the complexity and length of the study. At a minimum, rapport refers to respect, trust, and
positive regard between researcher and study participants that enhances openness and candor. Spradley
cautions that the mutual enjoyment typical of study participation does not necessarily translate into deep
friendship or affection. A researcher and study participant do not have to like one another to have rapport.

123
Presentation of Self
Qualitative researchers vary in how they present themselves information on their personal preferences as well
as the particular setting being entered. Erving Goffman (1961) adopted a marginal stance, detaching himself
as he observed asylums and other institutions. At the other end of the continuum, feminist researchers
advocate a research partnership in which roles are blurred and researchers are actively engaged in coproducing
findings with participants (Hesse-Biber, 2006). Most qualitative researchers present a “self” somewhere in
between these two ends of the continuum. Ben Henwood, a key member of the NYSS and NYRS teams
(before graduating to become an assistant professor), is a clean-cut young White man who was often jokingly
accused of being a cop by male participants. Yet his rapport with men and women alike was not hobbled by
his appearance once the conversations began.

The researcher is obliged to enter the field “with an open mind, not an empty head” (Fetterman, 1989, p. 11).
In other words, the researcher should be knowledgeable enough about the topic to prevent unnecessary
interruptions that disrupt the flow of the interview. In the NYRS, interviewers were trained to be familiar
with the slang words for drugs (e.g., crack cocaine is “rock,” synthetic marijuana is K2 or “spice,” and “benzos”
are anxiety medications abused on the streets). Similarly, the terms “SSI” and “rep payee” were common
parlance in interviews. These short-hand references to federal disability income and representative payee
status can be mentioned by the study participant without having to explain to the interviewer what they mean.
This leaves time for explorations with more depth and nuance, for example, questions about what it means to
have SSI as a steady source of income or to have someone else in control of one’s finances.

The optimal demeanor recommended for the qualitative researcher varies depending on personal style of
communication and engagement with others—creativity and intellectual curiosity should not be suppressed.
However, too much individuality or improvisation can detract from the real experts (the study participants).
Maintaining a sense of humor, a willingness to be wrong, and an eagerness to learn is a winning combination.

124
Researcher Self-Disclosure: How Much Is Enough? Too Much?
There are no prohibitions on a researcher sharing personal information in qualitative interviews; some is to be
expected as part of the give-and-take required to maintain rapport and trust. The overriding questions are:
Which “selves” should the researcher reveal? When is it appropriate to do this? When is it inappropriate?
Qualitative researchers generally adhere to the advice to be “truthful but vague” (Taylor & Bogdan, 1984).
One should never lie if asked a question, but one should also use discretion in how much to reveal. Weiss
(1994) suggests providing the basic “business card” information, but others assert that sharing personal
information encourages fuller disclosure by respondents and promotes greater equality (Gair, 2002).

Interviewees often ask questions that come out of the blue and require on-the-spot decisions on whether and
how to respond. Take the hypothetical example of a doctoral student researcher who is a licensed social
worker, a parent, an Army veteran, and an avid fan of bluegrass music. He is obliged to reveal his status as a
doctoral student right away because this is the requisite “self” conducting the study. If the study involves
potentially harmful subject matter, he will also have a caveat in the consent form regarding his legal status as a
mandated reporter of child or elder abuse. Whether he has the occasion to reveal his musical tastes or status as
a veteran and parent depends on the context.

A researcher’s disclosure of various selves and life experiences cannot follow a strict protocol but is usually
better confined to the before-and-after chat that accompanies each interview. During the interview or
observation period, such decisions should take into account a risk-benefit equation balancing rapport with the
potential for bias or intrusiveness. It can be entirely appropriate for the researcher to identify as being a
military veteran or a recovering addict, but this should not be done gratuitously or appear as attention-seeking
behavior.

125
Issues of Identity: Gender, Age, Race/Ethnicity, and Social Class
The dynamic interplay between researchers and their respondents—each affecting one another in
unpredictable ways—is a defining feature of qualitative inquiry. Feminist researchers have led the way in
discussing this as a methodological and moral issue (Fonow & Cook, 1991; Reinharz, 1992). Accepting rather
than condemning its existence, they have explored how intersubjectivity shapes and enhances the study
findings.

Whereas mutable characteristics such as dress and demeanor can be adjusted, relatively fixed traits such as sex,
age, race/ethnicity, and social class must also be considered. Socially constructed meanings associated with
such identities—and the intersectionality of such identities—often come into sharp focus in qualitative
research relationships. The consequences of such differences can be disastrous or negligible depending upon
how they are handled.

In addition to their demographic characteristics, researchers carry a number of acquired identities into the
field. These can be professional (student, professor, practitioner), personal (partner, parent, sibling), political
(feminist, climate change activist, gay rights advocate), and recreational (soccer fan, pianist, marathon runner).
How, if at all, these identities influence the study is largely a matter of context and appropriateness.

Sometimes the researcher has little in common with his or her respondents. In most instances, a successful
relationship need not be based on sameness (Manderson, Bennett, & Andajani-Sutjaho, 2006). Elliot Liebow
(1993) acknowledged stark differences in sex, age, race, religion, and social class when he began his study of
African American homeless women in Washington, D.C. Yet he was able to forge enduring relationships
with the women and write a moving portrayal of their lives. Along the same lines, feminist researchers have
noted that there are plenty of occasions when shared gender is not enough to instill trust and rapport. Perhaps
it is sufficient to say that disparities in the researcher–researched relationship deserve attention in the design
of qualitative research, whether that attention is concerned with training and supervision of the interviewers,
with ethical issues, or with how the findings will be safeguarded from bias.

126
Summary and Concluding Thoughts
Qualitative studies require careful thought to settle on the optimal design. Their contours are visible in a
thorough review of the literature—theoretical and empirical. In turn, the literature review sets the stage with
broad study questions and specific research questions containing key ideas and concepts. The selection of a
qualitative approach guides the design of the study and allows for an iterative back-and-forth rhythm between
data collection and analysis. This chapter has posed numerous questions whose answers form the backbone of
a study’s planning stage. Although each qualitative approach has its own procedures, all share in common the
need to transparently describe what lies ahead. Among design decisions, the choice of sampling strategy and
mode(s) of recruitment are pivotal.

Beginning a qualitative study requires careful planning amidst near certainty that things will not work out as
planned. Earning the trust of study participants and establishing rapport are necessary preconditions to
navigating through such uncertainty, but flexibility in design is also vital. Working without the protective
layers of distance and presumed neutrality, the qualitative researcher-as-instrument must maintain a level of
vigilance toward others as well as toward the self. One enters the field mindful not only of the physical
presentation of self but of the many identities that may come into play. Personal disclosure by the researcher is
a natural consequence of the give-and-take that often occurs before and after an interview.

127
Exercises
1. Choose a topic of interest and formulate one research question suitable for each of the types of
qualitative methods: ethnographic, grounded theory study, case study, and so on.
2. Using either the same topic you chose from #1 or a new one, plan and write out a research design for
one particular type of qualitative approach. Include answers to the “how many” questions as well as
which type of sampling strategy you will use.
3. Think of a “hard-to-reach” population, and identify a sampling strategy and recruitment technique that
would be most effective. Design a recruitment flyer that would likely pique their interest.
4. Go to your favorite journal and pick some articles that interest you. Read their literature review sections
and see if you can identify the argument being made.

128
Additional Readings
Crabtree, B. F., & Miller, W. L. (1999). Doing qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Creswell, J. W. (2007). Qualitative inquiry and research design (2nd ed.). Thousand Oaks, CA: Sage.

Gilbert, K. R. (Ed.). (2001). The emotional nature of qualitative research. Boca Raton, FL: CRC Press.

LeCompte, M. D., & Schensul, J. J. (1999). Designing and conducting ethnographic research (Ethnographer’s
toolkit, Vol. 1). Walnut Creek, CA: AltaMira Press.

Marshall, C., & Rossman, G. B. (2010). Designing qualitative research (5th ed.). Thousand Oaks, CA: Sage.

Maxwell, J. A. (2012). Qualitative research design: An interactive approach (3rd ed.). Thousand Oaks, CA: Sage.

Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

129
4 Ethical and Emotional Issues in Qualitative Research

As discussed in the Prologue to the previous chapter, ethical concerns deserve ongoing attention in a
qualitative study. Given the close quarters of most qualitative research relationships, boundaries can be
unknowingly crossed, sensitivities may be unwittingly offended, and losses of privacy can unintentionally
happen. Yet there are also reassuring aspects to qualitative methods coming from the same closeness and
relational attention. The balancing act is acknowledging the pitfalls without falling into them. We begin this
chapter by discussing the primary ethical concerns associated with research in general and then focus in on
how these are dealt with in qualitative studies.

130
Deception and Disclosure
Deception, concealing the nature of the study and the investigator’s role, is prohibited in virtually all research.
The history of research deception, from Holocaust medical experiments to the infamous Tuskegee study to
Stanley Milgram’s electric shock studies, makes its practice essentially indefensible. Of course, a high-risk
medical experiment has far more potential for harm than an observational study of behavior in busy public
spaces. Yet even the latter example is questionable given sensitivities about deception.

Three interrelated aspects of research deception are at issue in deciding the rare occasion when it is
permissible: (1) its necessity to carrying out the study; (2) its potential for harm; and (3) its intentionality.
With regard to the first of these, human subjects committees have clear guidelines restricting use of deception
to studies where the benefit outweighs the risk and the study could not be conducted otherwise. In these
instances—most of which are psychology experiments conducted with students—researchers are required to
put in place a number of safeguards such as debriefing subjects afterward. The potential for harm must be
minimal.

A number of landmark studies in the social sciences could not have been carried out without deception.
LaPiere’s (1934) study of racial and ethnic discrimination in the 1930s was based on his travels with a Chinese
couple to various hotels and restaurants around the United States. The proprietors’ discriminatory practices
were, unbeknownst to them, contrasted with their previous survey responses stating they did not practice
discrimination. If they had known the identity of LaPiere and his traveling companions, they likely would
have behaved differently.

LaPiere’s deception seems worth the risk, especially since the subjects’ identities were not revealed, and they
were observed behaving as they normally would. Similarly, a researcher interested in unobtrusively observing
panhandlers and pedestrians on a busy street can make a reasonable argument for deception as having little
potential for harm. To require informed consent would deny the naturalness on which the study depends. On
the other hand, observational studies in public or semipublic places can be out of line if they involve reports of
sensitive or stigmatized behavior. Laud Humphries’s (1970) landmark study of gay men’s behavior in public
restrooms provoked cries of outrage when his deceptive tactics became known. Although Humphries
countered that the study’s findings justified his use of deception, few would defend his actions today.

If the reasons for intentional deception are few and must be carefully justified, what about unintentional
deception? Human subjects committees would not countenance this on the face of things. However, in certain
types of qualitative research (e.g., ethnography), it is virtually impossible to notify every individual who might
cross the researcher’s path. Indeed, such instances call into question whether deception is happening at all.

Consider a hypothetical study of staff–patient interactions in a hospital emergency department. A good faith
effort requires that the gatekeepers (medical directors, administrators, and house staff) are notified in advance
and that formal sit-down interviews occur only after gaining informed consent. Yet one cannot give advance
notification to every individual passing through—every ambulance driver, family member, and janitor—by

131
telling them that they will be observed! Nor can one be expected to obtain formal consent every time an
impromptu question is asked. Promoting candor and transparency does not mean that participants have to be
told every detail of the study or that every person within sight of the researcher requires full notification—not
even the most stringent ethical guidelines require this.

132
Informed Consent
Because most qualitative research involves active, face-to-face engagement, informed consent is an “ongoing
and negotiated” process (Waldrop, 2004, p. 238). The basic elements of informed consent are

a brief description of the study and its procedures as they involve participants (approximate number of
interviews, duration of the study, etc.);
full identification of the researcher’s identity and of the sponsoring organization (if any), including an
address or telephone number for future contacts;
an assurance that participation is voluntary and the respondent has the right to withdraw at any time
without penalty or loss of services;
an assurance of strict confidentiality (which may be accompanied by two caveats: one regarding
mandated reporting by licensed professionals as required by state law and the other regarding the risk of
a breach in confidentiality from other focus group members); and
any risks or benefits associated with participation in the study. (Incentive payments are not considered
benefits but, rather, reimbursement.)

It is also necessary to get consent to audiotape interviews, along with assurances that participants may request
that all or part of such recordings be withdrawn from the study. Most researchers bring two copies of the
consent form (one for the researcher and one for the respondent to keep).

Special precautions are needed for studies involving members of vulnerable populations (e.g., pregnant
women, prisoners, institutionalized mentally disabled, and children). For studies of children and adolescents
under age 18, consent must be obtained from the parents and the child (children under age 12 may give verbal
assent). For vulnerable populations such as the frail elderly, consent may need to be obtained from a guardian
as well as from the respondent.

Signed consent may be waived under certain circumstances to protect the identities of vulnerable participants.
A researcher studying gay and lesbian youth at a community center where they are considered emancipated
would neither want nor need signed parental consent. The same would be true of undocumented immigrants
who have reasonable fears that disclosing their identity would jeopardize their status.

The researcher should obtain the permission of gatekeepers whose approval is necessary to carry out the study,
usually in the form of a signed letter. Such written permission is essential if their cooperation involves
assistance in recruitment. Gatekeepers may include agency directors, clinic supervisors, hospital
administrators, school principals, or the national health minister. To neglect this important task could cause
delays and even imperil a study. As discussed earlier, gaining informed consent for impromptu interviews
during field observation is not plausible; tacit consent is usually considered appropriate as long as the potential
respondent is free to refuse cooperation by simply walking away.

Obtaining formal consent has its price. Despite a requirement that consent forms be easily understood, many
researchers have witnessed the off-putting effects of asking participants to absorb two or more pages of

133
information cast in bureaucratic language and then put their signature on it. This distancing and formality has
to be overcome to set the stage for the freer expression to follow.

134
Coercion and “Deformed” Consent
The threat of coercion—whether heavy-handed or subtle—is a genuine concern in all research, especially in
studies of vulnerable populations. Researchers tend to occupy higher-ranking social positions than their study
participants and have institutional affiliations that can inspire feelings of coercion (not to mention real
coercion). Thus, even carefully obtained consent can become “deformed” consent. Human subjects
committees have extensive guidelines to safeguard against coercion, including consent form language designed
to assure prospective study participants that they are free to refuse participation and to withdraw from the
study at any time without any loss of services to which they are otherwise entitled.

For practitioner researchers, the potential for coercion becomes problematic when the study involves clients,
students, or coworkers who are familiar to them. For example, an administrator may want to conduct a study
in her agency, or a teacher may ask his students to participate in his research project. Interrupting a
professional relationship with one’s client, student, or coworker to ask for consent can appear (and be) coercive
even when handled sensitively. It is hard on the prospective respondents (who do not want to displease or who
may fear retribution), and it can be hard on the researcher (trying to shift between research and work roles).
Qualitative research is incompatible with the practitioner’s role when the two are carried out simultaneously
(Padgett, 1998). It is hard to see any satisfactory way to blend the two roles because the demands of being a
practitioner preclude the free, open-ended flow of information that is the essence of qualitative research.
Teachers and agency supervisors are similarly constrained when they consider studying their students or
subordinates.

Finally, the payment of incentives for study participation is a common practice if the researcher’s budget can
accommodate them (more on this topic a bit later in this chapter). However, it may be construed as coercive if
the amount is large enough to give the appearance of purchasing cooperation.

135
Confidentiality and Privacy
Qualitative researchers cannot offer the anonymity or safety in numbers that quantitative researchers can.
They must, however, provide virtually ironclad guarantees of confidentiality. This means that every effort is
made to ensure that the identities of participants are never revealed or linked to the information they provide
without their permission. Breaches of confidentiality—one of the utmost violations of trust—are undertaken
only in dire circumstances in which there are serious risks of harm to self or others, particularly children. For
licensed clinicians who are also researchers, mandated reporting is a legal requirement and must be so stated
on the consent form.

These precautionary measures represent a worst-case scenario that is extremely rare in qualitative research.
Although they encourage candor and openness, qualitative researchers are not usually a sounding board for
participants’ thoughts of harming others (even when such thoughts exist). The same is not necessarily true of
suicidal thoughts. In the NYRS, a few interviewees talked about suicidal ideation, usually as a thing of the
past or in the context of help seeking already undertaken. If a suicide attempt were probable and imminent,
interviewers were trained to return to the topic after the recording had stopped, to ask if the participant felt he
needed help and to offer to make a referral to a hotline or other resource.

Unlike quantitative research, a qualitative study runs a significant risk of breaching confidentiality in the
reporting of results. Pseudonyms are typically used and inconsequential facts changed to help prevent this
from happening (see Box 4.1 for a rare example of the problems this can cause). Such breaches are less likely
when the report relies on brief quotes or excerpts. More worrisome are vignettes or case studies in which
individuals’ life experiences are kept intact, and their identities are traceable by others who know them. The
more context and detail included in the report, the greater the risk. Moreover, gathering and storing visually
recognizable images (photographs or video) runs a much higher risk of exposure of one’s identity (or at least a
higher level of fear that this will happen). Compared to audiotaping, visual images are more identifying and
exposing. The combination of privacy laws and human subjects protections requires that signed releases and
informed consent be obtained from all persons in advance of photographing or videotaping them. Though
rarely considered, some study participants may want to be identified and credited, and this should be an
option if anticipated. In oral history research, for example, participants may want their names and faces to be
known because their identity is an important aspect of the historical record.

There is one other (albeit remote) risk of breaching the promise of confidentiality in qualitative research.
Unlike attorneys and physicians, researchers do not enjoy legal protection from demands that they disclose
information on illegal activities committed by study participants. Thus, an ethnographic study of drug dealers
could catch the attention of a local district attorney and inspire her to subpoena field notes and transcripts. If a
study’s participants are deemed vulnerable to such scrutiny, the researcher may wish to obtain a Federal
Certificate of Confidentiality (CoC) from the National Institutes of Health (available to all researchers
regardless of funding source). The CoC gives legal protection to participants for a specified period of time.
Thankfully, such dire outcomes are exceedingly rare. Prosecutors have better means (e.g., paid informants) to
track down illegal activities and are usually reluctant to incur the wrath of local universities or research

136
organizations. Of various threats to confidentiality, this is by far the least likely to occur.

137
Distress and Emotional Harm
Many qualitative interviews elicit intense discussions of painful life events such as divorce, death of a family
member, and domestic abuse. Common sense dictates that such topics are not introduced gratuitously; they
should be voluntarily brought up or, if necessary, inquired about carefully and empathically. Extremely
sensitive topics (incest, child abuse, suicidality) must be handled with the greatest care. When essential to the
study, emotion-laden topics are preferably left for the second interview.

Although human subjects committees often assume that talking about sensitive topics is a recipe for
psychological damage, emotional displays by respondents are not uncommon and are rarely cause for alarm. In
the NYRS, interviewers were trained to anticipate this, and they witnessed intense anger as well as mournful
sobs. Such episodes turned out to be infrequent and transient. Rather than resent the qualitative researcher for
eliciting such feelings, most participants remarked that the interview was an emotional catharsis for them, a
chance to express themselves in the presence of a nonjudgmental, sympathetic listener.

The researcher should make advance arrangements for referrals to professional counseling when emotional
responses are likely to occur. If a trained clinician, he should not provide this assistance directly (even if
respondents request it). If offered, referrals should be made after the interview has ended. It is important to
maintain this distinction between the data gathering part of an interview and the informal conversation and
sharing that take place when the recorder is turned off.

138
Incentives, Payback, and Maintaining Goodwill
Small monetary payments or other incentives encourage participation and partially compensate respondents
for their time. As mentioned earlier, funded research projects routinely include incentive payments in their
budget. The size of the payment depends on how much is being asked in terms of time and inconvenience.

This decision regarding how much to pay respondents is ethical as well as financial. If paying too little, the
incentive value is lost. But if paying too much (and especially if prospective respondents live in poverty), one
risks taking advantage of their poverty by purchasing their cooperation. Novice researchers usually consult
with more experienced colleagues to find out the current rates of compensation for study participation.
Payment of cash incentives may not be feasible for researchers who lack funding, especially students who are
themselves struggling financially. Doctoral students I have known have offered gift certificates to a coffee shop
or grocery or persuaded a local vendor to donate small items (tote bags, sundries, gift soaps, etc.).

Another form of compensation or payback takes place naturalistically during the study. For example, a
researcher who is spending time at a senior center interviewing staff might be able to offer an on-the-spot
tutorial on a computer software program to help them with an administrative task. In the NYRS, a study
participant with knee problems asked the study interviewer to accompany her on the long subway and bus ride
to her methadone clinic. The NYRS staff person agreed, and their conversation en route turned into a
valuable interview.

A final and often overlooked source of payback occurs after the study is completed—sharing the findings with
the participants. Respondents often want and deserve to see the results of a study in which they played a key
role, even if in the abbreviated form of an executive summary (more on this subject in Chapter 9).

139
Institutional Review Boards and Qualitative Research
A discussion of ethics in research would be incomplete without referring to the federally mandated regulatory
system that oversees adherence to standards of ethical conduct in the United States. Begun for the right
reasons—egregious abuses of human rights in biomedical experiments—the formal guidelines promulgated
under federal ruling 45 CFR 46 (the “Common Rule”) have raised a few vexing issues for qualitative
researchers. These include the wide latitude given to institutional review boards (IRBs), shifting definitions of
“research” and what thereby falls under their regulatory authority, and the methodological biases of these
guidelines in terms of their “fit” with nonbiomedical nonquantitative research. The distance between what is
prescribed by the federal Office for Human Research Protections (OHRP; http://www.hhs.gov/ohrp) and the
actual rules and restrictions implemented locally is a reflection of the interpretive freedom accorded IRBs.

Researchers in the social sciences have voiced concerns about IRBs since their inception in 1978. These center
on the excessive attention given to procedural issues and the tendency to view all research as suspect (a “guilty
until proven innocent” stance). In what has been an inevitable response to the expansion of their jurisdiction
and the rapid growth in empirical research, IRBs have come to rely on administrative staff to handle
applications, interpret federal guidelines, craft replies to researchers/applicants, and monitor ongoing
compliance (Fost & Levine, 2007). Being risk averse, local IRBs tend to delay or withhold approval when in
doubt. This is sensible on the face of it, but the definitions of risk and doubt tend to treat low-impact research
the same as more invasive and potentially harmful research. In my experience, being proactive in making one’s
case to the IRB brings dividends as well as being responsive and flexible when concerns are raised.

Since the federal stance regarding what constitutes research hinges on motivation or intent (i.e., whether one
plans to publish or disseminate the findings), virtually any investigation that involves human subjects would
fall under IRB jurisdiction. This has ramifications for the vast arena of social and health care services in which
management information systems and internal program evaluations can (and do) regularly verge into the
terrain of research. The primary deciding factors appear to be whether and when the program evaluator plans
to draw broader conclusions. What if the program evaluator starts out with local (nonresearch) intentions but
then changes her mind after preliminary analyses convince her that the findings should be used to generate
broader knowledge? Here, one encounters the blurry line between primary and secondary analyses. Data
collected without a priori informed consent may be viewed as a secondary analysis worthy of exempt status if it
contains no identifying information.

Many areas of IRB oversight remain open to varied interpretation regardless of the methodology being
employed. For example, do student research projects fall under the “normal educational practices” exemption
provided by federal guidelines? Is journalistic writing exempt from IRB review only when done by journalists?
If a study’s methods appear deeply flawed, should the IRB withhold approval given an apparent absence of
scientific worth? How is enforcement of noncompliance with IRB rules carried out equitably? Do researchers
who believe an IRB has overstepped its authority have avenues of appeal? Answers to these questions are
largely at the discretion of the researcher’s local IRB.

140
On the brighter side, most IRB staff members are ready and willing to help the researcher meet the standards
they are required to uphold. It helps to preemptively explain that qualitative methods are “low risk” even if
they cannot be described as “no risk.” Indeed, qualitative studies rarely exceed the typical IRB threshold of
“risks encountered in everyday life.” Concurrently, one takes care to put in all necessary safeguards to prevent
misunderstandings and delays.

141
Dealing With Moral Ambiguity and Risk
The moral ambiguity that surrounds naturalistic inquiry ensures that ethical dilemmas can arise at any time.
In-depth interviews can bring accounts of horrific behavior to the surface. Respondents may reveal aspects of
themselves that cause shock, anger, and feelings of exasperation. One NYRS participant robbed a credit union
to support her drug habit. Another related that he had once been a hit man for a drug lord and found killing
surprisingly easy. Jane Gilgun’s (1995) work with incest perpetrators of necessity reports on repugnant
behavior.

The intensity and trust unique to qualitative research allows respondents to feel safe enough to utter
despicable opinions or admit to illegal and morally reprehensible acts. The decision to intervene is rarely easy
and frequently has unforeseen consequences. Steven J. Taylor (1987) wrote of an extremely difficult situation
he encountered: physical and verbal abuse of developmentally disabled adults by the attendants in a residential
facility where he was conducting fieldwork. Although appalling, the attendants’ behavior triggered a moral
dilemma for Taylor when he reviewed his options for responding.

Intervening with the attendants might have inhibited their abusive behavior (at least in Taylor’s presence) but
at a cost of breaching confidentiality and losing rapport (thereby ending his inhibiting presence and the
study). Blowing the whistle and notifying the authorities (facility administrators, the police, or the media) was
made complicated by the fact that facility administrators knew about the abuse, tolerated it, and even covered
it up when confronted by family members. Taylor noted that pointing the finger at a few attendants was not
likely to prevent future abuse because it was so prevalent and tolerated. Blowing the whistle also comes at a
cost—breaching confidentiality and effectively ending the study.

Although acknowledging that there are occasions when intervening is worth it, Taylor ultimately decided to
take no immediate action against specific individuals (thereby protecting confidentiality). Instead, he carefully
documented his observations. After completing the study and writing about the prevalence of institutionalized
abuse in treatment of the mentally retarded he led a media campaign to expose the abuse and worked with
legal advocacy groups to draw attention to its prevalence. What at first seemed morally unacceptable—
continuing the study—gave him the commitment and knowledge to pursue these activities after the
conclusion of the study (Taylor, 1987).

The primary issue for conscientious qualitative researchers caught up in such situations is not whether to do
something, but what, how, and when to do it. The lesson of Taylor’s experience is that initial impulses to take
action should be weighed against foreseeable consequences. Put another way, the timing of taking action
should be calibrated to fit the immediacy of the threat and the likelihood that such action will produce the
desired result.

Moral ambiguity can happen in any study. Consider a few examples. A young woman with an ovarian tumor
says that she will not seek further treatment because she fears it will make her infertile. A college fraternity
member describes an initiation rite involving attempted rape. An immigrant mother insists on finding a

142
doctor to perform female circumcision on her young daughter. Least controllable are naturalistic go-with-the-
flow observations of behavior. During the NYRS, Ben Henwood was conducting a go-along interview with a
study participant in Harlem when a television news crew approached them and asked their opinions on the
new traffic signs installed by the city. Should Ben reject the crew’s overtures? Should he agree to the interview
along with the study participant (and risk appearing together on the nightly news)? As it happened, the study
participant withdrew from the encounter and waited patiently for Ben to politely extract himself (and Ben was
not on the news that night).

Ethnographic work such as shadowing interviews is especially prone to the unexpected. What if you are
shadow interviewing a participant and she demonstrates how she shoplifts items from a local pharmacy? What
if a participant invites you to the bar where he deejays on weeknights? (The latter occurred in the NYRS, and
the male interviewer readily agreed. It turned out to be a revealing introduction to the participant’s work
aspirations.)

143
Box 4.1 Moral Quandaries in an Ethnographic Study
A media firestorm broke out after a much-heralded ethnography was published by sociologist Alice Goffman (daughter of Erving
Goffman) in 2014. Based upon almost seven years of fieldwork in a poor neighborhood of Philadelphia, On the Run: Fugitive Life in
an American City (2014) was a meticulous and heart-felt report of young African American men and their families living under near-
constant police presence and frequent harassment. Poverty-stricken neighborhood residents also lived amidst drug-related violence,
and Goffman did not shy away from noting that local young men were responsible. She chose to live in the neighborhood and
became the roommate of two young men, one of whom (“Chuck”) was murdered by a rival gang member during the study.

Goffman’s book was well timed, as it struck a nerve already jangled from increasing attention to the racial biases attending police
killings of African American men and the phenomenon of mass incarceration (Alexander, 2010). Goffman’s empathy and eloquence,
not to mention the wrenching details of life on “6th Street,” lent the book an authenticity that garnered widespread acclaim in the
media. Then the criticism began.

Critics on the left complained that Goffman sensationalized African American violence, and critics on the right questioned why she
did not blame the perpetrators enough. Her living arrangements with male roommates were rather unusual, but it was her assertions
about certain police actions that got the most negative attention (Lubet, 2015). Adding fuel to the fire was her candid admission of
engaging in a dangerous, potentially criminal, act. Upset over Chuck’s murder, Goffman agreed to drive the car while an armed male
friend hunted for the killer. They did not succeed in finding the killer that night but not for lack of trying.

When confronted with accusations of factual inaccuracies and criminal conspiracy, Goffman insisted she was reporting what she saw,
heard, and did, her candor a necessary part of the reflexivity of modern ethnography as well as her deliberate alterations of facts to
“anonymize” her informants (Neyfakh, 2015). Goffman’s defenders (including her Princeton professors and her colleagues at the
University of Wisconsin) found the accusations of academic misconduct without merit, but this did little to stop the criticism.
Ultimately, research ethics and ethnographic methods were put on trial in the court of public opinion. Asked to back up her
assertions with field notes or other research data, Goffman replied that she had destroyed them to protect her study participants.
This revelation, along with the anonymizing of participants, was defended as standard practice in ethnographic research.

Some journalists and academic critics found this inability to fact-check or corroborate the study’s findings inexcusable and took the
opportunity to condemn ethnographers for promoting secrecy and lacking accountability (Neyfakh, 2015). At what point does
anonymizing move the dial from nonfiction to fiction? From empirical findings to memoir? Those familiar with the neighborhood—
including the residents who were tracked down and interviewed by a journalist (thus validating Goffman’s concerns about their
privacy)—stated that the book was a fair and accurate portrayal of their lives. Notwithstanding a few minor inaccuracies the larger
truth of the work appeared unassailable. However, methodological concerns about opacity and ethical concerns about researcher
boundary maintenance did not go away.

144
Ethical Issues in Community-Engaged Research
As described in Chapter 2, community-engaged research (CER) entails sensitive and respectful work in the
community where the study is being conducted. Despite an avowed commitment to improving lives and
environments, CER researchers may still go astray ethically. What if researchers who hire and train local
residents as co-researchers are accused of favoritism? Or what if community members resist being interviewed
by one of their own, fearing a loss of privacy and confidentiality? If observation is being carried out, how does
one ethically notify residents without undermining the naturalness required? Do CER researchers run the risk
of overidentifying with the cause such that they lose research distance? What happens when (not if but when)
researchers and community members disagree? How does the researcher overcome his or her status as
representing a powerful institution?

These questions might scare off some would-be CER researchers, but it is better to anticipate than react after
the fact. Connolly (2008) offered a cautionary tale of her community-engaged study with young women living
in a group home. She noted that recruiting some of them as co-researchers led to breaches of confidentiality
and near-violent encounters between the co-researchers and other female residents (Connolly, 2008). As with
so much of qualitative inquiry, the degree of trust and rapport enjoyed by researchers and community partners
can go a long way in ameliorating such problems.

For communities united by traditional cultural ties, researchers must pay heed to historical transgressions as
well as current circumstances. Many indigenous communities have taken steps to retain sovereignty over
research being conducted within their jurisdictions. In response to decades of exploitation, Native American
and First Nations tribal governments in the United States and Canada have codified strict guidelines to ensure
research ethics are observed. Guidelines specific to their concerns include collective as well as individual
consent; incorporation of indigenous knowledge; protection of cultural knowledge; and shared control over
data collection, analysis, and interpretation (Shore, Wong, Seifer, Grignon, & Gamble, 2008).

Human subjects committees tend to be unfamiliar with the principles of CER, especially those of collective
consent and shared control. Prominent among IRB concerns is the sharing of research data and potential
breaches of confidentiality. Many IRBs require that community partners undergo human subjects training and
certification, but this is a minimal approach to a much larger issue, that is, the messy realities of CER. As
often happens with nontraditional research methods, the onus is on the researcher to anticipate this and
educate the IRB using resources at hand.

And such resources have never been more available. The Community Campus Partnerships for Health
(CCPH) at the University of Washington has helped over 1,800 local organizations form nationally to
promote CBPR (Shore et al., 2008). As the list of resources at the end of Chapter 2 attests, virtually all
aspects of CBPR have been covered. As to research ethics and CBPR, there is a listserv specifically for this
purpose (https://mailman13.u.washington.edu/mailman/listinfo/ccph-ethics). In keeping with the values of
CER, a spirit of generosity and sharing permeates these online communities.

145
146
Ethical Issues in Cross-National Research
As qualitative researchers increasingly work across national boundaries, questions arise regarding IRB
jurisdictions at the foreign research site. Since Western researchers are typically affiliated with an academic or
other research institution, it is a foregone conclusion that they seek IRB approvals at home. The dilemma
comes from finding and honoring the relevant authorities where the research is to be carried out. In many
low- and middle-income (LMIC) countries, IRBs may be nonexistent or located in a distant city far from the
study location. Here, the need to find all gatekeepers and seek appropriate approvals is an urgent necessity.
The U.S. Office for Human Research Protections makes available an International Compilation of Human
Research Standards (ICHRS) that enumerates over 1,000 laws, regulations, and guidelines that govern human
subjects research in 113 countries. Developed for use by researchers, IRBs/Research Ethics Committees,
sponsors, and others who are involved in human subjects research around the world, this document can be
found at http://www.hhs.gov/ohrp/international/index.html.

The ubiquity of global research in health—especially drug testing and clinical trials—has prompted greater
attention to the needs and rights of research subjects, and rightly so. Abuses during such trials have been
egregious in some developing countries and have prompted international protests against pharmaceutical
companies’ lax procedures (Petryna, 2009). Even where there are strict laws and regulations in place, their
enforcement can be lackluster.

My experience in conducting research in Delhi, India (with the capable help of Indian-American field
assistant Prachi Priyam) was one of meeting the requirements of biomedical protocols unsuited to
ethnographic research (not to mention delays of a year or longer typical for approvals of research projects).
After several attempts at meeting the demands of a local research institute (which did not have jurisdiction but
would have conferred credibility), we gave up. The proverbial straw that broke the camel’s back came with a
requirement that we videotape the consent of homeless adults (“pavement dwellers”) rather than secure verbal
consent only. This requirement was an Indian government response to the discovery of deception used in a
clinical trial the year before and the deaths of some ill-informed subjects in the trial. Notwithstanding the
good intentions of this mandate, to carry this policy over to our study would have doomed it to failure.
Homeless men and women in Delhi are considered lawbreakers (begging is illegal, and most have to work in
the underground economy to survive), and they have to move around to avoid police arrest and harassment.
Not surprisingly, they are intensely suspicious of others’ motives. It was hard enough to gain trust, and a
videotaped consent would have been seen as a direct threat. In the end, we relied on our home institutional
approval and consulted regularly with the local nonprofit organization with which we had partnered.

147
Taking It Personally: Emotions and Qualitative Research
Interest in the emotion work of qualitative inquiry has grown as qualitative researchers are urged to become
more self-aware about their work (Blix & Wettergren, 2014; Dickson-Swift, James, Kippen, & Liamputtong,
2009; Johnson & Clarke, 2003). Influenced by a now-classic study of emotion management among flight
attendants (Hochschild, 1983), research interest in emotional engagement encompasses more than the
inevitably human reactions to intimate information shared in confidence. As noted by Blix and Wettergren
(2014), gaining rapport and pleasing gatekeepers as well as ongoing maintenance of these relational ties are
also part of emotion work. Ceding control to participants leaves the researcher in limbo at times and unable to
foresee all problems. More peripheral but nonetheless affected are transcribers and data analysts (Etherington,
2007). It is not uncommon for a transcriber to inquire about a participant’s well-being after having listened to
him or her relate a painful story.

Contending with expectations of professionalism elides the messiness of going where the participants are and
staying there as long as possible. Dickson-Swift and colleagues explored the dimensions of emotion work
among qualitative researchers using grounded theory interviews. Some reported crying along with the
interviewee, others held it in until later, and still others had few emotional reactions. The gendered aspect of
emotion work was in evidence, that is, women were more likely to acknowledge and display emotions, men
more likely to suppress them (Dickson-Swift et al., 2009). Gair (2002) provides a moving account of her
dissertation research on adoptive mothering, her feminist approach and shared experiences affording a degree
of self-disclosure and closeness that required constant monitoring on her part. The job of managing one’s
emotions while eliciting fraught information is not one to be taken lightly, yet it is often viewed as beneath
the dignity of a researcher.

Researchers studying a topic close to personal experience run an additional risk of becoming emotionally
enmeshed as the study proceeds. And, the ethical complications multiply when the researcher chooses a topic
and a setting that are close to home. Humphrey (2012) conducted a 4-year case study within her school of
social work that brought forth a number of insider ethical dilemmas related to her status as a social worker,
researcher, and former teacher. Assuming the role of humble learner places the qualitative researcher at odds
with the power dynamic of most research; emotional displays only further this distance from “science” and add
to the disorientation. Similarities to psychotherapy are inescapable. Ullman (2014) was mindful of this as she
interviewed therapists working with sexual assault survivors and kept a “meta” log of her own reactions to the
clinicians as well as the subject matter.

Debriefing with others helps enormously, especially when done on a regular basis. In the NYRS, we
scheduled weekly meetings to talk about interviews and observations, including their emotional impact. We
also visited a local bar every few weeks to drink a few beers and discuss nonresearch aspects of our lives.
Nevertheless, interviewers struggled with the seemingly unending accounts of loss, violence, sexual abuse, and
addiction. How does one walk away unscathed from a woman’s story of being raped repeatedly and
impregnated by her father? Or a man who witnessed his brother beheaded in a horrific auto accident?

148
The investigator working alone can seek out other qualitative researchers for debriefing as long as everyone
safeguards the confidentiality of their data (more on peer debriefing in Chapter 8). A thesis advisor or
research mentor can also provide helpful advice on how to stay on course emotionally. Keeping a private
journal is a favorite among qualitative researchers as it provides the opportunity to “let it all hang out” and—
an added benefit—get accustomed to writing and composing one’s thoughts about the study.

Emotional stress is not the only risk a qualitative researcher may encounter. Although extremely rare, threats
to safety may come from respondents, especially those with violent histories. Sexual come-ons and innuendo
can also occur. In her study of divorcing couples, Riessman (1990) found herself deflecting sexual advances by
male respondents. Such events can be discomfiting, especially when the interview takes place in the
respondent’s home or a private location where the researcher feels vulnerable. Our NYRS protocols called for
in-home interviews whenever possible, and we initially forbade female interviewers from going to a male
participant’s apartment—but this rule fell by the wayside as interviewers insisted they felt safe (we had no
problems as a result). Nevertheless, a rule of thumb in these situations is to plan ahead whenever possible, be
acutely aware of one’s surroundings, and to maintain professional poise (and leave the premises) whenever the
discomfort level becomes intolerable.

149
Socially Responsible Research as Ethical Research
Social responsibility as an ethical stance includes being sensitive to diversity—gender, social class, age,
ethnicity, and sexual orientation—but it encompasses much more. Professional organizations, as evidenced in
the National Association of Social Workers (NASW) Code of Ethics (http://www.naswdc.org/pubs/code)
urge their members to pursue research while being mindful of the potential for harm. The NASW Code of
Ethics also promotes social justice as fundamental to research as well as practice and policy.

When it comes to social responsibility toward research participants, qualitative researchers have an advantage
because they are obliged to “go where the respondents are.” (Action-oriented researchers go further by putting
social justice front and center.) Beyond concern for individuals, socially responsible research implies taking the
larger structural context into consideration when interpreting and understanding the data. A study of persons
with schizophrenia might, for example, conclude that they prefer to be isolated if the focus is narrowly on
what they say. Considerations of stigma and social exclusion bring an enlarged and more realistic perspective
on their social isolation.

The incorporation of broader socioeconomic and cultural perspectives may seem to conflict with a focus on
the intricate textures of people’s lives seen from the inside (the emic perspective) rather than from the outside
(the etic perspective). In fact, some qualitative approaches are less open to addressing structural issues—
narrative and phenomenological analyses come to mind—whereas others such as critical theory and feminism
are avowedly structural in focusing on inequality and its consequences. The majority of qualitative studies,
however, occupy a middle ground in which the researcher may choose to incorporate structural concerns at
some (or all) stages of the study.

Socially responsible research does not mean presenting a one-sided portrait that leaves out the less savory
aspects of respondents’ lives. Qualitative researchers are not investigative reporters digging for dirt, but they
are also not obliged to produce an uplifting portrayal devoid of authenticity. This delicate balance between
accuracy and sensitivity to respondents’ needs affects studies of the despicable, the heroic, and the everyday
people in between.

150
Summary and Concluding Thoughts
Ethical issues are omnipresent in qualitative inquiry. As with all research in the United States, qualitative
studies must adhere to federal guidelines ensuring voluntary informed consent; freedom from deception,
coercion, or emotional harm; and protection of confidentiality and privacy. However, many gray areas remain
when it comes to interpreting these guidelines and addressing ethical issues lying outside of IRB jurisdiction.

The intensity and duration of the qualitative research relationship, in combination with its lack of strict
neutrality, ensure that boundary maintenance requires constant vigilance. Notwithstanding good intentions,
researchers who encounter morally ambiguous situations risk inappropriate responses such as getting too
involved, showing disapproval, or letting their enthusiasm verge into coercion. The vast majority of qualitative
studies poses a risk for harm no more severe than those encountered in everyday life. If properly observant of
ethical guidelines, qualitative researchers can take comfort in the fact that their studies are likely to provide a
satisfying and memorable experience for respondents.

151
Exercises
1. Go to a university institutional review board website, and take the tutorial on human subjects
protections. What are the basic elements of ethical research conduct?
2. Draft a sample consent form that includes the essential protections of human subjects for studying each
of the following groups: (1) hospice patients dying of cancer; (2) parents of children with autism; and (3)
adolescents ages 12 to 16 who smoke cigarettes.
3. Consider the possibility of conducting ethnographic research in a busy pediatric asthma clinic serving a
poor inner city population. How would you approach informed consent since you will be visiting for
prolonged periods and at different times but cannot possibly request informed consent from everyone
who will be visiting the clinic?
4. Go to the NIH website explaining the federal Certificate of Confidentiality
(http://grants.nih.gov/grants/policy/coc). Discuss what types of study populations are most likely to
need this extra protection.

152
Additional Readings
Christians, C. G. (2003). Ethics and politics in qualitative research. In N. K. Denzin & Y. S. Lincoln (Eds.),
The landscape of qualitative research: Theories and issues (2nd ed., pp. 208–244). Thousand Oaks, CA: Sage.

Dickson-Swift, V., James, E. L., & Liamputong, P. (2008). Undertaking sensitive research in the health and
social sciences: Managing boundaries, emotions and risks. Cambridge, MA: Cambridge University Press.

Gilbert, K. R. (Ed.). (2001). The emotional nature of qualitative research. Boca Raton, FL: CRC Press.

Guillemin, M., & Gillam, L. (2004). Ethics, reflexivity, and “ethically important moments” in research.
Qualitative Inquiry, 10, 261–280.

Hammersley, M., & Traianou, A. (2012). Ethics in qualitative research: Controversies and contexts. London,
UK: Sage.

Kilty, J., Felices-Luna, M., & Fabian, S. (2015). Demarginalizing voices: Commitment, emotion, and action in
qualitative research. Vancouver, BC: UBC Press.

Malone, R. E., Yerger, V. E., McGruder, C., & Froelicher, E. (2006). “It’s like Tuskegee in reverse”: A case
study of ethical tensions in institutional review board review of community-based participatory research.
American Journal of Public Health, 96, 1914–1919.

Miller, T., Birch, M., Mauthner, M., & Jessop J (Eds.). (2012). Ethics in qualitative research (2nd ed.).
London, UK: Sage.

Morse, J. M. (2007). Ethics in action: Ethical principles for doing qualitative research. Qualitative Health
Research, 17(8), 1003–1005.

Thorne, S. (1998). Ethical and representational issues in qualitative secondary analysis. Qualitative Health
Research, 8(4), 547–555.

153
5 Data Collection Observation, Interviewing, and Use of
Documents

Of the three basic types of qualitative data (interview, observation, and documents), interviews have assumed a
dominance that has relegated the other two to distant second and third places. The enthusiastic adoption of
qualitative methods by talk-based professions such as teaching, nursing, and social work no doubt contributed
to this dominance. Talking and dialogue have intrinsic appeal due to their familiarity in everyday lives
(Atkinson & Silverman, 1997)—the same cannot be said of participant observation or the analysis of
documents. Nevertheless, all three sources of data play uniquely different and complementary roles. Although
not every source of data is appropriate for any given study, each should be given serious consideration.

154
Observation in Qualitative Studies
Imagine trying to study a community garden project, a methadone clinic, or a senior center using only verbal
description. The visual aspects of these settings are surely of interest, not to mention the naturalistic behavior
taking place among the actors. Many actors are unaware of goings-on in their midst or have spotty recall (and
those who have more to hide usually hide it more). There is no satisfactory substitute for being there in person
in real time.

A doctoral student I worked with, Justine McGovern, had worked closely with Alzheimer’s patients in her
practice and this piqued her interest in how married couples navigated this difficult transition in its early
stages. Initially proposing to do interviews with the couple and individually, Justine added a participant
observation component to better understand the dynamics of their relationships and the ways they
communicated with one another. This led to numerous trips to their homes, to local diners, to church
services, and even a bowling alley. The study also added a documents component when a husband or wife
shared family photos, artwork, poems, diary entries, and letters. Individual interviews allowed both members
of the couple—including the spouse with early-stage Alzheimer’s—to open up about their challenges in
adjusting to the disease and the joint (or linked) interviews brought out memories of their lives together. A
discourse or narrative analysis could have been an alternative study option for the latter.

It is helpful to distinguish between observation as a general mode of qualitative data collection and participant
observation, the centerpiece of ethnographic fieldwork. While the latter assumes the primacy of immersion
and interaction within a specific environment, the former is less site-specific and often used in conjunction
with interviewing (thus being less intense and time consuming). Even minimal amounts of participation
require careful consideration given the potential for reactivity (changes due to the researcher’s presence).
Think of a continuum ranging from full participation (living amid the study population for a prolonged period
and carrying out daily activities much as they do) to unobtrusive observation with little or no interaction. The
risks at the high-involvement end of the continuum include researcher fatigue and losing sight of the
researcher role. At the low level of involvement, avoiding interaction means foregoing a valuable means of
learning from those being observed.

Most common is a midlevel of participation that is selective and contextual. If the study is site-centered,
attendance at carefully sampled events—staff meetings, street festivals, political protests and so on—can be
useful in documenting the range of activities. If the study is person-centered, getting permission to shadow
study participants on their daily rounds of activities or on selected outings gives added value to what is
otherwise entirely dependent on the participants’ verbal accounts. Also known as “go-along” interviews
(Kusenbach, 2003), these encounters afford the opportunity to learn through observing and interviewing
simultaneously. Sampling of events may be representative (e.g., varying by time of day or types of participants)
or purposive (e.g., attending case conferences reserved for the most troubled clients or shadowing a study
participant on the first of the month after the disability check has arrived).

Here are a few examples of additional information we obtained in the New York Recovery Study (NYRS)

155
when we shadowed study participants (SPs) on their routine errands:

SP pointed out a local bodega where the kindly owner allows him to buy food and sundries on credit.
SP has to take two buses and a subway to get to her methadone program (over an hour each way).
SP has a friendly relationship with his local pharmacist, which makes getting his medications much
easier.
SP visits a city park daily to write in his diary and relax (the park is a well-known hangout for marijuana
dealers).

Each of these outings required preparation, that is, selecting the best day to meet, avoiding bad weather,
aiming for typicality (in one such incident, the NYRS interviewer spent an afternoon with the participant
watching daytime television game shows).

Shadowing or go-along interviews have their own challenges. First is the lack of privacy. Presumably such
encounters involve public spaces and for obvious reasons sensitive topics and behaviors need to be avoided.
The running commentary of participants during shadowing can be burdensome if they feel forced, and it can
attract unwanted attention from (or awkward conversations with) bystanders. Gill and colleagues recommend
planning for meal and restroom breaks and taking a spiral notebook (that folds open) along with a cellphone
for ad hoc audio recording (Gill, Barbour & Dean, 2014). As with all observation, committing to memory
interesting information is the best way to ensure getting it down. Deciding what and how to record, attending
to what the participant is saying and doing, and expecting the unexpected makes a shadow interview an
occasion for constant awareness.

156
“Doing” Observation: The Best (and Only True) Way to Learn
We are all participant observers in a way—nothing less is required when entering an unfamiliar situation. Yet
qualitative research demands a more disciplined and nonjudgmental form of observation than the necessarily
self-interested and selective observations one makes in daily life. The best approach is to begin by casting a
wide net then moving to more focused observation as the study’s aims and major themes begin to crystallize in
the data. All of the senses are involved—sounds and smells can convey important meanings and
environmental cues. Ethnographers usually seek out one or more key informants (knowledgeable individuals
who can supply valuable information). Although the term informant is a regrettable holdover from earlier
days, it connotes what is being sought because the individual is not asked to talk about personal experiences.

To illustrate the unique powers of observation with all of the senses, consider a typical homeless men’s shelter.
In photographs, it can appear relatively benign, for example, rows of beds with men sitting or lying on them.
But a few visits for field observation will bring the sensory experience of the shelter to full life. Pervasive theft,
drug use, and physical intimidation are the norms of life in the public shelter system, as recounted by residents
and often barely concealed from visitors. The sensory stimuli add an extra dimension or two. The smells—
body odor (few working showers), poor sanitation (nonfunctioning toilets), greasy cooking, and powerful
chemical disinfectants—can be overpowering. However, the noise and lack of privacy are often the most
disturbing aspects of shelter life—the cacophony of sounds from dozens (or hundreds) of men crowded into
close quarters, cursing and shouting (including guards and staff), some in the throes of drug-induced or
mental psychoses. Usually required to leave the premises for the day, weary occupants return at night for rest
but get little. Residents’ accounts of municipal shelter life are often dismissed as exaggerated or flimsy excuses
for preferring the freedom of the streets. But researchers’ first-hand observation more often than not validates
these portrayals of homeless shelters as multisensory hazards to health and well-being.

157
Recording Observational Data

Observation protocols . . . cannot be treated as faithful reproductions or unproblematic summaries

of what is experienced, but should be seen, rather, for what they are: texts written by authors, using
their available linguistic resources, to give a meaningful summary of their observations and
recollections after the event. (Luders, 2004, p. 228)

Field notes are the necessary if imperfect representation of what is experienced during observation (Emerson,
Fretz, & Shaw, 2011). The quality is improved significantly by adhering to two practices. First, field notes
should be taken either in real time or very soon thereafter because memory erodes dramatically within the first
24 hours. Second, it is important to avoid the clouding or distorting effects of “filters,” whether personal
predilections or other allegiances. This can happen by referring to a man and a woman arguing on the
sidewalk as a “couple” without knowing what their relationship actually is. It can also occur when using
psychological terms such as denial when referring to a respondent’s reluctance to address a sensitive issue. No
observer is a bias-free instrument, but attending to this fact is the requisite first step.

Much has been written addressing the “why, what, and how” questions of field note taking. Answering “why”
is the easiest to do: Logging one’s observations produces invaluable raw data that are archived for analysis. It
can also lead to more focused follow-up via interviewing or additional observation. The “what” of field note
taking depends on the topic of interest as no one can presume to record everything. General suggestions
regarding what to look for in the initial phase include physical space, actors, behaviors, interactions,
relationships, and affect or emotions (Lofland & Snow, 2005). If one is observing a bounded physical space, it
is useful to draw a map or floor plan to add a spatial dimension to what is happening. An agency where staff
share cramped cubicles will have a different feel from one where private workspace is the norm.

The “how” of field note taking requires a good deal of flexibility, sensitivity to the situation, and practice. It is
easier to take notes (or speak into a recorder or smartphone) in busy public spaces where one can remain
relatively inconspicuous. But there are times when writing notes would intrude on the natural course of events
or provide cause for offense. For these occasions, it is best to take mental notes—committing to memory as
much as possible—followed by brief jottings. (Retreating to the bathroom is a favorite technique for
accomplishing this.) At the end of the day, one can write in greater detail. Box 5.1 offers an example of field
notes from the NYRS.

158
 Box 5.1 Field Notes From a Client Meeting at a Transitional Housing
Program in the NYRS
Below are excerpts from field notes that have been slightly altered to protect confidentiality:

J. (African American woman) who is one of the program supervisors was on stage speaking with concern as security cameras at the
program’s building entrance showed clients smoking. J said that there is to be no loitering and panhandling in front of the program
building or the clients’ apartment buildings. She had to shout since the air conditioning unit was loud, and there was no microphone.
She repeated multiple times to the clients that it was important to be a good representative for the agency. “It’s important to send a
positive message to your neighbors. Every time you walk out of your apartment, you’re representing the agency.” She continued by
telling the clients that many of the buildings have a “no loitering” sign posted in front. “If you’re just standing there, you’re loitering.”
She asked the clients, “What does this say to your neighbors? Your neighbors will have a negative outlook on you and the agency.”

One client raised his hand to ask “How do people know that we’re in [program name]?” J. hesitated for a moment, then said “Your
neighbors don’t know unless you choose to tell them.” She then quickly turned to the next client with his hand up. This client (who
happens to be in our study) was quite agitated and stated that the neighbors know because the counselors keep shouting at the door
“Counselor! Counselor!” when they arrive for home visits. J. told the client that the case management team will have to discuss this,
then transitioned to the next topic on the agenda.

Lofland and Lofland (1995, pp. 89–90) provide a number of helpful hints for field note taking:

1. Aim for the concrete and specific in describing behaviors and events. At the beginning, try to avoid any
inferences, whether your own or volunteered by others.
2. Try to distinguish between the different types or levels of observational data based on their proximity to
the event being observed. First-order data such as verbatim accounts are recorded either during, or
immediately after, the period of observation. The second level of data involves paraphrasing
conversations and less certain recall after observation has taken place. Ideas regarding new directions and
inferences are further along the continuum of removal from the event and are recorded in analytic
memos.
3. Record observations of yourself—your impressions, feelings, and concerns. You can bracket this
information in your field notes or log it separately in a diary. Keeping a running commentary of personal
reactions and feelings serves two related purposes. First, it is an outlet—a place to unload the inevitably
human reactions to prolonged contact with others. Second, it provides a means of identifying personal
biases and devising ways to manage them.
4. Strive for balance—don’t let yourself become lost in a forest of minutiae, yet don’t lose the tendency for
compulsiveness that motivates the best field observers. Even small amounts of time spent in the field
translate into lengthy notes. The ratio is around 6 to 1—six hours of recording for every hour of
observation.

The average field notebook is not scintillating reading—it will appear wordy and tedious to an outside reader.
At times, the burden to record as much as possible can be onerous; most ethnographers have at least one story
of hiding from an informant to avoid yet another encounter that will need to be noted. After all, a 15-minute
conversation might turn into an hour or more of write-up.

159
160
Uses of Video and Photography
Video and photography are ancillary means of observation and also stand-alone methods of data collection.
Film and video have long been used in ethnography, pioneered by Margaret Mead and Gregory Bateson
during their fieldwork in New Guinea in the 1930s. Ethnographic films such as the 1920s classic Nanook of the
North and Frederick Wiseman’s Titicut Follies in 1967 chronicled the decline of traditional Inuit culture on
the one hand and the horrors of insane asylums on the other. Visual anthropology and visual sociology have
become two of the faster growing subfields in their respective disciplines and the Journal of Video Ethnography
was launched in 2015. The ubiquity of smartphones makes videotaping a readily available companion to field
notes. Two promising uses of photography are photo-voice projects and photo elicitation interviews.

Photovoice (PV) techniques, which originated in public health, offer an appealing option that enables study
participants to control the camera and tell the story (Wang & Burris, 1997). Influenced by the emancipatory
writings of Paolo Freire in combination with feminist theory, PV, like community-based participatory
research (CBPR), merges research with community empowerment. Its implementation, which involves
collaboration between researchers and community representatives, is a means for individuals and communities
to document their strengths as well as their concerns in order to effect positive change (Wang, Morrel-
Samuels, Hutchinson, Bell, & Pestronk, 2004). The targets of visual depiction might be sources of local
pollution that endanger health, community violence, the lack of green spaces, and so on.

Internet-mediated photo sharing makes PV possible without face-to-face contact (Yi-Frazier et al., 2015).
Instagram, a photo-sharing mobile phone application, has 300 million monthly active users, of whom 70%
reside outside the United States (Instagram.com/press). The possibilities for research with adolescents,
persons lacking mobility, and cross-national populations are many and varied. Yi-Frazier and colleagues
(2015) used Instagram-mediated PV with adolescents diagnosed with Type I diabetes; their shared photos
included cartoons and humor as well as self-care tips.

PV projects in health promotion and community empowerment have provided valuable opportunities to
democratize research practices and achieve collective action (Cabassa et al., 2013). The data include interviews
and photographs as expressions of a communal voice, and the end products include community forums and
advocacy as well as traditional research findings. Similar to CBPR, PV entails lengthy time and resource
commitments and subordination of one’s own research goals to those of community members.

Photo-elicitation interviews (PEI) are typically one-on-one rather than community-based like PV, but both
share a focus on the visual image as the pivot point. With origins in anthropology, PEI is a valuable means of
eliciting thoughts and memories that verbal-only interviews frequently overlook (Collier & Collier, 1986;
Harper, 2012; Pink, 2012; Rose, 2007). As richly informative as verbal interviews can be, photographs bring
forth new dimensions of color, composition, and creativity. When respondent-driven (rather than researcher-
controlled), photos not only document people, places, and things but evoke all manner of feelings.

161
162
 Box 5.2 Photo-Elicitation in the NYRS
In the NYRS, we turned to PEI as a means to engage participants to produce their own visual data. We had some trepidation and
warnings from others—the cameras would “disappear” or be returned broken, release forms would not be completed, photo taking in
questionable venues would get someone into trouble, and so on. With the exception of one camera lost, none of these things
happened.

Lacking time and resources to conduct PEI with all study participants, we purposively chose 18 who appeared capable of and
interested in PEI. In keeping with our interest in recovering lives after histories of mental illness, homelessness, and substance abuse,
study participants were given minimal direction beyond brief instruction on use of the digital camera and obtaining signed releases
from individuals they photographed. Participants were asked to take up to 18 photographs that depicted their lives, both negative
and positive, and to return in 2 weeks for the PEI. Upon the participant’s arrival, the interviewer took the camera to a local
pharmacy, downloaded and printed the photos, then asked the participant to arrange and describe each photo and what it meant to
him or her (the PEI was audiotaped and transcribed verbatim with the photos embedded in the transcript). Participants kept the
hard copies of the photos and were asked to repeat the PEI 12 months later, at which time they could keep the camera.

The PEIs proved to be revelatory in more ways than we could imagine. We were struck by the willingness of participants to divulge
earlier traumas (unmentioned in previous interviews), to celebrate their sobriety, and to reflect on urban life and space—all through
the medium of visual images. The multidimensionality of the experience was immediately apparent. First was the way they carefully
arranged and displayed the photographs when sitting down at a table with the interviewer. Second, the content of the photographs
could be categorized (we chose to distinguish photos that were “documenting,” “representing,” and “evoking”). Third, we found that
participants took two basic approaches to PEI overall (we termed these “a slice of life” and “then vs. now”) (Padgett, Smith, Derejko,
Henwood, & Tiderington, 2013). In further analyses, we found that the photographs and accompanying narratives constituted
“identity work” in which participants gave voice to their attempts to forge positive identities, to leave behind negative identities, and
to have a deeper appreciation of their environmental amenities such as parks, community murals, and public art (Tran Smith,
Padgett, Choy Brown, & Henwood, 2015).

The description of PEI in Box 5.2 opens the door to wider discussions of the role of visual data in qualitative
research. As PEI provides an occasion for sharing thoughts and experiences with an interviewer, meaning
making through narrativizing is taking place (Holstein & Gubrium, 1995). Similarly, some photographs
provided straightforward description (e.g., a park where the participant liked to go), but others invited a
constructivist approach. For example, a participant might take the opportunity to create meaning from
seemingly mundane items (e.g., a photo of a new pair of sneakers signified greater mobility but also having a
fixed address to order items online for home delivery). Some participants used metaphors in their meaning
making, for example, photographing an open road under blue skies to denote a more positive future.

Mapping is also useful as a source of visual data. Long popular with quantitative researchers who use
Geographic Information Systems (GIS) and social indicator data to track everything from epidemics to crime
rates, mapping locates important information in space and time. This permits identifying high concentrations
of phenomena (e.g., suicide clusters) as well as their absence (e.g., food deserts). In line with the value placed
on subjectivity and user perspectives, qualitative researchers use participatory mapping to locate “activity spaces”
(Vallée, Cadot, Roustit, Parizot, & Chauvin, 2011), or they might use drawings of concentric circles to plot a
participant’s inner and outer social networks. In the NYRS, we geo-coded the locations of the PEI
photographs on a map (with the help of Google) and classified them to identify where participants went for
leisure purposes as well as service use.

163
Of course, geo-coding and mapping can raise understandable fears of surveillance; participants need to feel
comfortable that their privacy is protected. With proper attention to ethical protections (such as leaving out
identities and keeping locational specifications under lock and key), mapping, diagrams, and floor plans lend a
unique strength to studies where space and place are important.

164
In-Depth Interviewing
High-quality interviews are the linchpin of success for qualitative studies. Yet in-depth interviewing has
become the subject of recent debates in light of the trend toward greater reflexivity and epistemological
awareness (Atkinson & Silverman, 1997; Holstein & Gubrium, 1995). The issues range from macro (the
consequences of living in an “interview society”) to micro (the dialogic act itself). Atkinson and Silverman
(1997) have led the way in calling out the Oprah-like elevation of interviewing as a form of personal salvation.
The “truth” of an interview is in doubt, as it is a second-hand account crafted for the occasion (Alvesson,
2011).

Much effort has been expended on training the qualitative interviewer to go beneath the artifice of crafted
stories. The goal is to be attuned yet unbiased, empathic but not overly solicitous, judicious but also vigilant in
the use of probes (more about probes later in this chapter). Such is the focus of this book and most textbooks
on qualitative interviewing, and rightly so. But closer examination of the dance that is a qualitative interview
also requires some thought about what we are being told. In my experience, interviewees rarely lie outright;
they are more likely to leave out than fabricate. Still, this issue bears further consideration given the
pervasiveness of interviewing in qualitative research.

165
Critiques of Interviewing: Realism and Authenticity
The heavy reliance on interview data that has emerged over the past few decades has sparked debates about
truthfulness (respondents lying), incompleteness (respondents’ omissions), and the well-known gap between
what people say and what they do (Hammersley & Gomm, 2008). Bourdieu’s (1999) writing on “false
collusive objectification” inspired Yanos and Hopper (2008) to critique qualitative interviews as too easily
falling into pretense where the interviewee gives a “rehearsed account,” and the interviewer willingly obliges.
This collusion can stem from the interviewer’s satisfaction in hearing familiar talk (e.g., interviewees speaking
in the jargon of self-help) or from the postmodern belief that there is no truthful rendition to be sought.

For example, redemption stories are not uncommon among individuals accustomed to pleasing service
providers and others in power; anyone can be forgiven for having a tendency to prettify, obscure, or talk in
ways pleasing to the interviewer. In the New York Services Study (NYSS), we heard catchphrases from drug
rehab lingo (“I was sick and tired of being sick and tired”) that were distancing from the gritty realities of
recovery. And we stoically but skeptically accepted a participant’s insistence that he no longer smoked
marijuana (the sense of smell being available to us) because knowing that ‘truth’ was not essential.

The problem comes from the interviewer’s inability (or unwillingness) to get beneath the surface, all with the
complicity of the interviewee. Critiques of interview-only studies have intensified with the valorization of
interviewees’ subjective accounts as “personal truth.” In addition, the postmodern-inspired shift to treating the
interview as a dialogic performance rather than a window into human experience renders moot the question of
whether respondents are lying, obfuscating, or contradicting themselves (Atkinson & Coffey, 2002;
Hammersley, 2008). As noted by Yanos and Hopper (2008), “lack of authenticity is apparently of little
concern when one is focused purely on the interview as a vehicle of identity construction.” (p. 233). Relatedly,
being a “good” interviewer is less of a concern since one need only encourage participants to speak without
interruption. Indeed, some forms of conversation analysis avoid interviews altogether in favor of naturally
occurring speech (Speer, 2002).

According to Hammersley (2008), qualitative interviews have been used in four related but distinct ways: (1)
as witness accounts of events and experiences; (2) as self-reflection by the interviewee; (3) as indirect sources
of information the researcher can draw on to make inferences about the interviewee; and (4) as purely dialogic
or discursive events. Hammersley’s concern lies with the discounting of the first three of these. Such critiques
come from a realist perspective concerned with balancing the search for authenticity with respect for
subjectivity.

Researchers in social work, nursing, and public health have a vested interest in making their research noticed
in venues beyond the interview room or academic conference. Here is an example from NYSS interviews
where having the truth mattered and hopefully informed better practice. In participants’ life history
interviews, they told us that their initiation into substance abuse came well before their first psychiatric
hospitalization, usually in the early teen years. Most grew up in close proximity to drug and alcohol abuse, if
not in their immediate family, then in the neighborhood. In the literature on persons with dual diagnosis,

166
researchers focused on current circumstances and surmised that drug and alcohol abuse was a response to the
mental illness, a form of “self-medication” (Khantzian, 1985). The truth of participants’ accounts mattered to
us since the onset of substance abuse long predated psychiatric problems, thereby challenging received
wisdom. Treatment providers could benefit from knowing that psychiatric medication may not substitute for
this other “medication,” the latter instigated by powerful environmental cues, not psychotic voices (Henwood
& Padgett, 2007).

Sometimes it is important to know what is real (or at least understood as real) and other times not so much.
Narrative approaches focus on how something is said more than what is said (although content is not excluded
from scrutiny). In phenomenological research, the lived experience is sought and appreciated rather than
interrogated. Case studies, grounded theory (GT), ethnography, and action-based approaches treat interviews
primarily as sources of data.

Participants’ subjectivity should not be tampered with unless there is good reason. Sometimes subjectivity is
the point, for example, a participant who relates his traumatic experience of sexual assault. At other times,
decisions have to be made regarding whether to pursue follow-up questions to get at the accuracy of an
assertion or to clarify what the interviewee meant to say. This could be unnecessary when other sources of data
—observation, documents, and collateral interviews—can be used to fill in the blanks. In the absence of such
information (or when there are discrepancies), doubts do not lead to interrogation and a breach of hard-won
trust. If a participant’s account seems to be entirely misleading or concocted, the interview may need to be
excluded from the study altogether.

The issue of telling the truth in qualitative interviewing is not one to get mired down in since respectful
engagement almost always produces authenticity. Keeping this in mind, the remainder of this chapter is
devoted to maximizing the richness and depth of qualitative interviews. Such interviews vary along several
dimensions, including methodology (ethnographic interviews differ from phenomenological interviews),
format (one-on-one vs. group), medium (in-person vs. electronic), and degree of structure. We begin with
informal field interviews then proceed to formal in-depth interviewing in its many varieties.

167
Informal Field Interviews
The informal give-and-take that is a critical part of observation is an ad hoc form of interviewing. Field
interviews are context driven and rarely amenable to advance preparation although the observer has the study’s
goals in mind when posing questions. Informal field interviews are not small talk and chitchat—they are a
targeted effort to get needed information. They usually entail a brief exchange (e.g., “Can you tell me how sex
education is taught in this school?”). Such impromptu queries serve as the interstitial glue that binds long
hours of observation with more formal in-depth interviews. Without them, the informational gaps would
make in-depth interviews much longer and more tedious. Informal field interviews may be audio recorded
(with permission), but they more often end up in field notes They do not require formal consent, but
sensitivity to the person and the context is essential.

In the NYRS, ethnographic site visits consisted of attending case manager meetings and client “check days”
(when clients came to the program offices to pick up their weekly allowances deducted from their disability
checks). Informal interviews were common in these situations, as the observer would pose impromptu
questions to staff and clients. Remaining careful not to let those questions intrude or distract from the
business at hand meant that some had to wait until the timing was right. Study team members were often
invited to the clinical supervisor’s office to debrief and ask questions for this purpose.

168
Focus Group Interviews
Focus group methods originated in sociology (Merton, Fiske, & Kendall, 1956) and were developed for use in
marketing and polling so that small groups of unrelated individuals could be brought together to discuss a new
product or political candidate (Krueger, 1994). The size of a focus group should be large enough to generate
diversity of opinions but small enough to allow everyone to share in the discussion—five to seven participants
is about right, but as few as three can work (Morgan, 1997).

As originally conceived, the focus group comprised persons from similar backgrounds who did not know each
other—familiarity can lead to more habitual ways of interacting and inhibit fresh opinions from emerging.
Focus groups that include persons from different levels of a status hierarchy are considered problematic
because the subordinates have understandable concerns about the repercussions of being candid. In practice,
both of these assumptions are violated. In a typical social service agency, for example, it would be near
impossible to convene focus groups of staff members who do not know one another. Less common but also
possible are focus groups in which some supervisory staff are included, and the topic is benign enough to allow
this to take place without threatening underlings.

Focus group interviewing requires a moderator who asks open-ended questions, but the degree of direction
and structure can vary depending on how narrow or broad is the topic of inquiry. In addition to the need to be
sensitive, flexible, and empathic, the moderator must avoid certain pitfalls common to group situations
(Fontana & Frey, 1994). These include domination by one person or a clique and silent (or sullen) group
members. Poor group leaders—those who dominate the discussion or are too passive—can make focus groups
a lost cause. Even skilled moderators can find it difficult to guide members away from internal dissension or
divisive tactics.

The logistics of focus groups require special considerations, both methodological and ethical (see Box 5.3 for
an example). These days, scheduling a time and place to meet can be a real challenge—early evenings often
work best but not always. Inducements such as refreshments and incentives almost always help with
attendance and comfort level.

The optimal structure of the focus group format is where turn taking is frequent, and the moderator exerts
control only to ensure a smooth and informative discussion. If at all possible, group moderators should not be
tasked with taking notes because their attention needs to be on the group members and what they are saying.
MacGregor, Rodger, Cummings, and Leschied (2006) conducted focus groups with foster parents that were
co-facilitated by a parent coordinator and university professor, with a graduate research assistant in charge of
recording and taking notes.

With respect to data collection, audiotaping is increasingly used, but some researchers and/or group
participants prefer to have either a scribe or note-taker only. Either way, complications may arise when trying
to distinguish speakers without revealing their identity—some moderators ask members to adopt a pseudonym
or ID number at the outset and use it each time they speak.

169
As with any method of data collection, focus groups have limitations. An ethical problem can occur when a
group member breaches confidentiality. The researcher has no control over this rare but unhappy event and
can only warn prospective group members of this possibility in the consent form. If a group member decides
to withdraw from the study, the researcher may be obliged to expunge his or her statements from the
transcript.

170
 Box 5.3 Preparing for a Focus Group Interview: The Need to
Multitask
In the final phase of the NYRS, we conducted focus groups (FGs) at supportive housing programs in Washington, DC,
Philadelphia, Burlington (Vermont), and New York City. Compared to a one-on-one interview, the planning and preparation were
considerably more laborious. Our study design required that we interview three groups of stakeholders at each program—consumers
(clients), front-line staff, and supervisory staff. Thus, we asked each program to work with us on scheduling the groups and
recruiting attendees. While we asked our contact person at each program to avoid “cherry picking” clients for the groups (the
programs were small enough that this was not possible with the staff focus groups), we could not control how this was done without
interfering with program operations. The programs generally preferred to do all three FGs on the same day so this meant putting all
of the preparations in place well in advance and getting plenty of rest before the appointed day. Two NYRS team members were
involved in each FG (one to facilitate and the other to take notes). Led by NYRS Project Director Bikki Tran Smith (with the able
assistance of Emmy Tiderington and Mimi Choy-Brown), the preparations involved buying healthy snacks and drinks, making
copies of consent forms and interview guides, and putting incentive payments ($20) in envelopes along with receipts to be signed for
bookkeeping purposes. At the mercy of the program’s ability to accommodate our presence and assemble FG attendees, we traded
several e-mails and telephone calls in advance.

Focus groups developed and matured outside of the evolving traditions of qualitative methodologies, and their
economy and convenience sometimes lead to less-than-rigorous methods. Agar and McDonald (1995) raise
concerns that “a few hours with a few groups guarantees only that the ‘quality’ in qualitative will go the way of
fast food” (p. 78). The format of a focus group tends to discourage going into deeper or sensitive areas
(Morgan, 1997). In this era of busy schedules, it has become more difficult to find an amenable time for all
attendees, and no-shows are a problem as well as late arrivals and early leavers. The FG setting can make a
difference—an office location with a laid-back ambience might be easier to instill rapport but less organized to
support FGs. Similarly, a well-run but rigidly supervised setting can meet the organizational needs of FGs but
at the expense of attendees’ comfort in speaking up.

Despite these potential drawbacks, focus group interviewing can bring clear advantages, including savings in
time and resources and the elicitation of insights stimulated by the group dynamic. It is a particularly good fit
for studies in organizations or communities where there is a web of social networks already in place. Group
interviews are useful in conducting needs assessments, in studies of organizational change, and in community-
based studies.

171
Individual Interviews
Ideally, in-depth interviewing is scheduled in advance, takes place in a private setting conducive to trust and
candor, and is preceded by careful preparation. Of course, there is immense variation in how an interview is
planned and how it unfolds. At its best, it is a one-sided conversation. Box 5.4 offers some general tips for
interviewing in qualitative studies.

172
 Box 5.4 Tips for Qualitative Interviewing
1. Familiarize yourself with your questions as much as possible in advance.
2. Ask follow-up questions based on what you are hearing; use the participants’ words when possible.
3. Avoid leading questions.
4. Explore issues but don’t interrogate.
5. Feel free to not understand (but don’t come across as patronizingly clueless).
6. Try not to rely on questions about feelings—let feelings emerge naturally, and then ask about them.
7. Encourage participants to share anecdotes and specific experiences—avoid generalities.
8. Monitor personal disclosure; emphasize rapport-building over drawing attention to yourself.
9. Don’t interrupt or try to control the interchange.
10. Accept pauses as natural; break the silence only if the participant seems stuck.
11. Feel free to laugh and appreciate humor.
12. Avoid becoming informal and “knowing” in follow-up interviews.
13. Take notes for follow-up questions, but don’t let them become distracting.
14. Remember that everyone has a “bad interview day.”

In-depth interviews follow different rhythms of directivity based on the type of method being employed.
Some studies derive their interpretative power from unbroken narratives in which the way a person tells the
story is of interest as well as the content of what is being said. Morse (2015) refers to this as unstructured
interviewing. Examples include phenomenological, life history, and narrative interviews. In these instances,
probes or follow-up questions are minimized to avoid interruptions.

Other interview methods (e.g., grounded theory) are more dialogic and semi-structured. The “minimally
structured” interview is one in which a set of questions is developed in advance, but there is flexibility in their
sequencing and whether all or used or not. At the more prearranged end of this semi-structured continuum—
but still far from the rapid back-and-forth of a quantitative survey—are interviews in which the same set of
open-ended questions is asked of every participant in the same sequence.

Interviews in the NYRS differed significantly across its two phases, the first phase elicited free-flowing life
histories, and the second phase was a semi-structured inquiry into participants’ experiences with drugs and
alcohol, their social networks, and their use of mental health and other services. Both phases placed a
premium on open-ended questions, but Phase 2 included “must-ask” questions to ensure that certain domains
were addressed uniformly with all respondents.

Most in-depth qualitative interviews are minimally structured, but they are not full improvisation. Typically,
the interview guide reveals the study’s key domains, that is, classes of information from which a variety of
questions could be asked, some prepared in advance and others elicited through probes. Deciding on what
constitutes must-have information is important because anything not routinely elicited will emerge only if
volunteered or resulting from an ad hoc probe. In the NYRS, we chose not to directly ask about traumatic
experiences because we wanted to respect participants’ privacy and their emotional state. Nevertheless, such
information did come up, and in startlingly graphic detail (Padgett, Smith, Henwood & Tiderington, 2012).
This is another example of how sensitivity and rapport in qualitative inquiry can make it possible to obtain

173
important information without eliciting it directly.

Although traditionally preferring a one-on-one format, some qualitative researchers have noted the utility of
two interviewers, one to ask the questions and the other to observe, take notes, and probe if the occasion arises
(Alvesson, 2011). This could work in situations where respondents are comfortable with a two-on-one
interview, but the need should be made clear in advance since obtaining rapport is obviously complicated in a
triad as opposed to a dyad.

174
Linked Interviews
A variant on the individual interview arises in qualitative studies where persons are of interest because they are
linked together in some way, for example, a married couple, a mentor-mentee relationship, parent and child,
and so forth. Unlike a focus group interview, linked interviews are intended to explore personal experiences of
the individuals alone and in conjunction with each other, and the interpersonal relationship or linkage is an
important focus of the study. Although most often involving a duo, linked interviews could ostensibly be done
with three or four members of a family or other small interconnected group. Of course, linked interviews with
more than two individuals increase the need to balance the connections and multiple facets of their
relationships. At some point, the number of individuals and perspectives may attenuate what can be learned
through linked interviews.

Decisions about sequencing of the interviews arise early on. Should each individual be interviewed separately
before the joint interview or after? Renee Spencer’s (2006) qualitative research on mentoring relationships
between adults and youth introduced the complications of race/ethnic and social class differences to linked
interviews. Her interviewing approach followed a sequence of first meeting the pair to describe the study and
obtain consent, then interviewing each person individually, and finally reconvening the mentor and mentee for
a joint interview (taking 1 ½ to 2 hours total). Such an approach enables the researcher to get acquainted with
both participants but also relate to each as an individual.

175
Interviewing Children and Other Vulnerable Populations
As mentioned in Chapter 4, human subjects committees have federally mandated protections for vulnerable
populations. Because qualitative interviews with prisoners and pregnant women are more a matter of avoiding
coercion than using specialized interview techniques, I will focus here on minor children and persons who are
mentally impaired in some way.

Interviewing children—alone or with a parent—has challenges related to their vulnerability (to parental
authority and the researcher’s presence). As well, young children lack the verbal abilities, life experiences, and
insights that adults bring to interviews. Older children and adolescents have more verbal capabilities but may
be uncooperative or have short attention spans. Developmental stage and maturity can make the difference.

Despite ethical and logistical challenges, qualitative interviews with children can be done (Alderson &
Morrow, 2004), especially when video, photography, and observation are also used (Barker & Weller, 2003).
Berguno and colleagues (2004) interviewed 42 children 8 to 13 years of age about their experiences of bullying
and feelings of loneliness at school—the interviews lasted on average 20 minutes. Highly appropriate for
younger age groups is the use of role-playing and games (Mufti, Towell, & Cartright, 2015). Similarly,
documents—diaries, drawings, essays, and photographs—can be useful sources of data to supplement brief
interviews.

Persons with mental disability may have organic impairment due to brain injury, or they may have a serious
mental illness such as schizophrenia. For those on the severe end of the disablement continuum, qualitative
interviewing is not feasible. However, the decision to exclude participants with mental impairment should be
made carefully because stereotypical assumptions can interfere with their right to be heard (not to mention the
researcher’s need to learn from them). Several years ago, I met with a doctoral student interested in homeless
mentally ill women. This student planned to interview these women’s case managers. When asked why she
did not talk to the women directly, she insisted that they did not have the verbal or cognitive abilities to
engage in qualitative interviews. This decision denied her access to the true experts.

The underlying tension here is between acknowledging the abilities of children and mentally impaired persons
(including their right to be research participants) and protecting them from abuses of power related to the
conduct of research. The ability to remain sensitive to individual vulnerabilities tips the balance toward
inclusiveness.

176
Key Informant, Elite, and Expert Interviews
Valuable for what and who they know, key informants help pave the way to more knowledgeable observation
and interviewing. They may be drawn from the ranks of community leaders but just as often are individuals
who have insider knowledge of all rungs of local society, that is, a retired schoolteacher. Unable to speak the
native language or communicate without a translator, traditional ethnographers depended heavily on key
informants who might also function as translators. The use of key informants brings vital information forth,
but it does not substitute for interviews with a broad variety of individuals about their lives and experiences.

Elite or expert interviews may target highly regarded practitioners (e.g., a famous heart surgeon), policymakers
(e.g., a state legislator in charge of prison appropriations), or other public figures (e.g., a leading advocate for
children’s rights). These individuals add a top-down perspective that would otherwise be missed without their
participation. For a qualitative study of foster care in which I was involved, key stakeholder groups included
youths but also family court judges and leading child advocates (Freundlich, Avery, & Padgett, 2007). Such
interviews require special planning and foresight. Busy professionals and civic leaders are difficult to schedule
time with. They have little to gain and often fear they have something to lose from talking to a researcher.
Questions usually need to be tailored individually to ensure maximum use of time and draw on the unique
perspective of the interviewee.

177
Developing the Interview Guide
The interview guide starts with the basics at the top of the first page: the date, time, location of the interview,
and name or initials of the interviewer. Demographic characteristics (age, sex, race, education, etc.) are elicited
at the close of the interview—an alert interviewer only asks questions that remain unanswered in the course of
the interview.

Qualitative interview guides consist of open-ended questions, usually the fewer the better. The spontaneity
and flexibility that make qualitative interviews succeed come from probes, some of which can be anticipated
and others are spontaneous. If the scope of inquiry is broad, the interview guide may be organized around
domains or topical areas. The following (see Box 5.5) are two examples of domains from the NYRS and
sample questions and probes associated with each. In the interview guide, we put the probes in italics to
remind the interviewer that they were to be asked only if the interviewee did not cover that information
spontaneously. Note that one of the probes also comes with a prompt and that “SP” means “study participant.”

178
 Box 5.5 Examples of Domains from the NYRS Interview Guide
Domain: Entering the Program

Sample Question #1: How did you get to this program?

Probe for:

Source of referral
Degree of choice in entering the program
If SP was homeless prior to coming to program

Domain: Social Networks

Sample Question #1: Who, if anybody, can you count on the most to help when you need it?

Probe for:

Relationship to SP
Type of help given. Prompt: financial help, child care, food, cash, other

Sample Question #2: Is there anyone who counts on you for help?

Probe for:

Relationship to participant
Type of help given. Prompt: financial help, child care, food, cash, other

Sample Question #2 (“Is there anyone who counts on you for help?”) was at first considered a probe but
subsequently earned its way to full question status. The usual rule is to upgrade a probe to a question if it is
“must-have” information that cannot be left to discretion. As it happened, this question not only revealed
neglected facets of participants’ lives but also helped enhance rapport by signaling that the interviewers were
not focused one-sidedly on problems and woes.

When developing an interview guide, pay close attention to the wording to avoid confusion or leading
questions. Consider a scenario in which you are interested in whether and how participants plan for the
future. As with any domain, there are many ways to ask about this. The worst from a qualitative point of view
would be a question along the lines of “What are your goals for the next 5 years?” Not much better would be
“What are your goals for the future?” The tacit message here is one that many individuals cannot relate to,
especially if they are having personal problems and find goal-setting language to be judgmental and intrusive.
In the NYRS, we settled on easing into the subject by asking “What is the next step for you?” The “next step”
question turned out to be a winner for us, giving participants a chance to reflect on the near or far term as they
saw fit and to talk about their hopes as well as concrete plans. In the final NYRS interview, we inserted a
probe after the “next step” question, asking “And how about the longer-term future—where do you see
yourself?”

Charmaz (2014) urges GT interviewers to focus on processes in their questions. Asking “what led up to___?”
“what happened next?” “what was going on in your life then?” or “how does the current situation compare to
earlier?” captures the action surrounding key events. In later rounds of GT interviewing, the focus gradually

179
shifts toward theoretical concerns identified earlier. Consider a hypothetical study of recovering from
addiction in which “natural recovery” appears frequently in participants’ accounts. The idea that one might
summon the courage and strength to do this gradually and without treatment could prompt more questions
and probes to flesh out what natural recovery entails.

Interview guides may be standard for all participants, tailored uniquely to each person, or a combination of
both. Questions should allow for contingencies such that the interviewer may skip a question that is not
relevant or already discussed (e.g., leaving out a set of questions on substance abuse if the interviewee says she
has never used drugs or alcohol). Sequencing of questions is very important. Avoid starting out with
demographic and other factual questions that give the wrong impression. Also avoid starting out with highly
sensitive questions—save these for later, when rapport is established.

The best opening gambit is to ask an open-ended but non-threatening question, one that engages the study
participant without putting him on the spot. The best closing gambit is to ask the participant if he or she
would like to add anything or has any feedback on the interview experience. This debriefing shows respect as
well as concern. If emotions run high at times, they should resolve by the end, or more debriefing should be
allowed. Turning off the recorder and easing out of interview mode into small talk brings a more pleasant
sense of closure. In the unlikely event that a participant is still upset and there is reason to believe things could
worsen, the interviewer should follow study protocols regarding making “the offer” to refer to counseling.

180
Elicitation Techniques in Qualitative Interviews
Qualitative researchers may use elicitation techniques in which participants are asked to answer a question by
listing their responses or reacting to a stimulus such as a photograph or list of terms. Elicitation techniques are
designed to prompt responses within a designated format. Originating in anthropology where the goal is to
understand cultural beliefs and practices, these techniques include photographic elicitation interviews
(described earlier in this chapter), free-listing, and pile-sorting (Weller & Romney, 1988).

Free-listing entails asking the interviewee to identify the items belonging to a domain, for example, street
names for illegal drugs, reasons for avoiding jury duty, what are low-fat snack foods. Probes or prompts may
be necessary to keep the list-making going, but the assumption is that the interviewee is in charge of what
goes in the list and what does not. Pile-sorting—often taking place after free-listing has produced the set of
responses—starts with a set of prompts, and the interviewee is asked to arrange them in some order (Bernard
& Gravlee, 2014).

Perhaps not surprisingly, by producing quasi-standardized responses, these techniques can become the basis
for quantification (Borgatti, 1994). Free-lists require categorizing and sorting before calculating frequencies,
and pile sorts open the door to counting what is put into the domain versus what is left out as well as how the
items are ranked or ordered. Matrices and tables are handy devices for doing this. Focusing on the most
frequently listed or higher-ranked items allows the researcher to identify salience (Rugg & McGeorge, 2005).

In the NYRS, we wanted to understand participants’ life priorities in their final interview. We arrived at card
sorting after trying a free-list approach that failed (participants shrugged or came up with vague answers). The
card sorting method involved compiling a list of possible responses to the question “what is important in your
life?” Each of 12 answers drawn from the mental health recovery literature was placed on a laminated 3x5
card, and the participant was asked to choose answers and then rank each according to its importance (Choy-
Brown, Padgett, Tran Smith, & Tiderington, 2015). The cards included physical health, mental health,
romantic partner, job, housing, and so on. We were pleasantly surprised by how much participants enjoyed
the exercise and having control over the cards. Some probing was necessary, for example, “why did you leave
out X?,” “what makes X your top priority?,” but the card-sort exercise turned out to be a successful technique
(Choy-Brown et al., 2015).

181
Interview Guides for Second (and Third) Interviews
Since multiple interviews with each participant bring depth as well as the opportunity to clarify and expand, a
question arises: what should a follow-up interview guide include? This is where the flexibility and iterative
nature of qualitative studies comes into play, especially for phenomenological and GT studies. The multiple
interviews of phenomenological studies have the goal of going deeper into lived experience. In GT,
subsequent interviews may be used to test out and expand on initial hunches (from the first interview or from
interviews with other participants). This use of theoretical sampling to generate new interviews is a classic GT
technique.

In practice, the second interview is often the completion of the first interview where time ran out (as it often
does). Of course, new questions can and should be added after the first interview—many interviewers
welcome the second chance to clarify a major point or to venture into a new direction as participants’ wariness
melts away. Not surprisingly, it is often a challenge to specify the content of the second interview in advance,
and such specificity is not always desirable. If one’s advisor, human subjects committee, or other person in
authority requests a copy of the second interview guide, this can be done with the proviso that additional
questions may be called for. Whenever possible, the concept of saturation may be invoked to explain why the
number of interviews may be larger (or smaller) than projected.

182
A Few Guidelines for Starting Out
A qualitative interview is conversational without being a conversation, a distinction that is not always easy to
recognize or put into practice. In addition to the tips listed in Box 5.4, a few helpful hints come in handy.
First, clarify your stance as much as possible, situating it along a continuum of structure and directedness.
Although study participants are encouraged to see the interviewer as a curious, uninformed learner, the
interviewer provides gentle guidance and directs the flow of the interview. It may be a partnership, but the
roles remain different and complementary.

Second, pilot test the interview guide on a few individuals (preferably drawn from the population of interest).
Giving the interview guide a trial run almost always leads to changes, whether paring down the number of
questions or minor tweaking of wording. Third, if the study is a team effort, build in plenty of time for
training interviewers, mixing didactic instruction with role-playing exercises. If time permits, transcribe the
training interviews, and use the transcripts as teaching devices to point out strengths and weaknesses in
technique. If time is short, use the early interviews for the same purpose. Fourth, plan for the location of the
interviews, attempting when possible to leave this up to the participant. Privacy, comfort, and safety are
paramount.

Last but not least, it is always a good idea to start and end the interview with small talk when the recorder is
not running. This helps produce a smoother beginning, particularly after going over the details of the consent
form. At the end of the interview, casual repartee leaves participants with a sense of being appreciated, a
chance to pose questions or make comments, and to lay the groundwork for future interviews and contacts.

183
Conducting the Interview: The Importance of Probes
The ideal interview is one that creates space and comfort for the interviewee to speak and the interviewer to
improvise. Some of the most valuable information in qualitative interviews comes from probes, of which some
are planned, but most are spontaneous. The interviewer must remain alert to subtle cues. Probes are critical
for getting beyond rehearsed accounts and prefabricated renditions. Box 5.6 gives an example of the
importance of probes from a previous study in which I was involved.

184
 Box 5.6 The Importance of Probes Both Planned and Spontaneous:
“Air Theory” and Breast Cancer
In an earlier study in which I was involved (The Harlem Mammogram Study), we were interested in why some African American
women neglected to follow up with diagnostic testing after receiving notice of an abnormal mammogram. In the interviews, we
asked about their beliefs regarding what caused breast cancer and its spread throughout the body. Almost in passing, a few women
mentioned a firm conviction that “air” is partly or fully responsible. An alert interviewer felt comfortable enough to probe further.
Respondents obliged by explaining that opening up the body during surgery exposed dormant cancer cells to the air and precipitated
their growth and spread to other areas of the body. During our regular interviewer debriefings, we agreed that this unforeseen
information should be gently probed for because it could shed new light on why women do not follow up on recommendations for
surgical biopsies and other diagnostic services. We preferred contextual probes rather than direct questioning; the latter might be
perceived as too leading or presumptive.

Subsequent interviews and probes revealed varied forms of “air theory” attributions among several of the women, all sharing a general
concern with the dangers of ambient air during surgical procedures. A couple of years later, in a surprising twist of fate, this
potentially harmful “folk belief” turned out to have more credence than previously thought. In short, it received the biomedical seal
of approval when surgical researchers at Columbia University reported that airborne pathogens may explain the greater recurrence of
cancer after open-incision surgery compared to closed-incision (laparoscopic) surgery!

For some interviewees, one question will release the floodgates, but most qualitative interviews rely on in-the-
moment probes to obtain the depth and richness desired. Impromptu probes can be used to

go deeper (“Can you tell me more about . . . ?”);

go back (“Earlier you mentioned__________ , please tell me. . . .”);
clarify (“And were you homeless when you were arrested?”);
steer (“That’s very interesting, but can we return to . . . ?”); or
contrast (“How would you compare your experiences in foster care to living with your adoptive family?”).

Each of these has a role depending on what is being said and what is being sought. The “go deeper” probe
usually opens the door wider than the others. Here, the interviewer must avoid gratuitous questions that can
appear voyeuristic if not part of the study’s goals and the interview’s flow. Probes about sexual functioning can
be appropriate if part of a discussion about the side effects of antipsychotic medications but inappropriate if
brought up with young adults talking about their dating experiences.

The “steering” probe is used judiciously so as not to interrupt the flow. Respondents are known to go off on
tangents, some of which might later yield nuggets of insight while others are yawn inducing. It is not unusual
for an interviewer to listen attentively through lengthy harangues about an ungrateful child, boasts of sexual
conquests, and arguments in favor of the legalization of drugs. Steering probes come into play when time
grows short and content becomes thin. Probes can be built into the interview guide, but they also emerge
spontaneously. A qualitative study that relies on multiple interviewers must ensure that such improvisation
does not introduce “solo performances” by interviewers keen on making their stylistic mark.

Less commonly used are prompts (i.e., suggested options offered when a question is vague, but a checklist is
not appropriate). When asking participants if they have used any health services in the past 6 months, one
might prompt them by mentioning “clinic visits, emergency rooms, and the like” to give the respondent some

185
idea of what “health services” refers to. Obviously, care needs to be taken in using prompts because they may
put words in a respondent’s mouth.

186
After the Interview
The interview feedback form (IFF) is a place to log observations of the participant (she seemed hostile,
intoxicated, distracted, eager to please, etc.) and about the setting (the participant’s home was immaculately
kept; the interview room was interrupted by staff, etc.). It is also valuable to jot down any concerns about the
participant and ideas to follow up. Box 5.7 provides an example of an IFF from the NYRS with some facts
changed to shield the participant’s identity. This one is relatively brief, other IFFs in the study ran up to three
pages in length.

Interviewer observations capture many things unsaid, including tone of voice (sarcasm, sadness, light-
heartedness), speech impediments, facial expressions (grimaces, winks, smiles), body language, and the
ambience of the setting (background noise, unsanitary smells, interruptions). All provide a feeling for the
context that is missing from the transcript if not otherwise noted. What if the respondents’ words contradict
their actions and circumstances? In these instances, it is best not to point out the discrepancy or force a
confrontation over what version of events is the “truth.” Prolonged engagement with respondents will likely
bring this out in time (although a briefer relationship might leave the researcher guessing). In any event, such
observations of participant discrepancy can be noted in the IFF.

187
Box 5.7 An Example of Interviewer Feedback From the NYRS
NYRS Interview Feedback Form

(Note: This form should be filled out soon after the interview and brought to the weekly debriefing meeting. If possible, a typed
transcript should be included.)

Participant ID#: 21X M/F:M Age: 36

Race/Ethnicity: Af-Am (Caribbean)  Date of Interview: 02/05/12

Site of Interview: NYSS office  Interviewer: BH

Time Begun: 10:30 a.m.  Time Ended: 12 p.m.

1. Note study participant (SP)’s demeanor and mood during the interview (anxious, impatient, relaxed, angry, etc.).

SP seemed relaxed, friendly, and open to talking. When I arrived, his roommate was home who is also a participant in our study, who
greeted me with surprise and told our new SP that I’m a “great guy.” Once his roommate left, I sat down with the SP who was open to
answering all my questions. Overall he was not difficult to follow but is from Trinidad and has an accent that along with his
enunciation made certain phrases hard to understand
2. Note observations of SP’s posture or nonverbal behavior. (Remember to include gestures, expressions, and utterances such as
laughter or crying in transcript!)

SP sat at his kitchen table with his laptop computer in front of him. He sat mostly straight in his chair and usually looked at his
computer screen as he was speaking. He was not very expressive but did greet me with a smile that I saw on occasion throughout the
interview (such as when he played some of his rap music for me that is available on ITunes).
3. If the interview took place at SP’s home, note physical setting, orderliness, personal artifacts such as photos, and so on.

The apartment was well lit, warm, and seemed clean. It was sparsely decorated with nothing on the wall except a subway map. SP
commented that he cleans the place regularly, but that his roommate does not do a good job of cleaning.
4. Summarize briefly SP’s history of drug/alcohol use and current status.

SP said he was never heavily into drugs or alcohol because he witnessed how much damage those things caused people he knew. He said
he occasionally used marijuana.
5. Note any concerns you have about the quality of the information given by SP, including sobriety and mental state as well as
your own reaction to the SP.

No concerns.

188
Matching Interviewers to Respondents: The Effects of Age, Gender, Race,
and Other Characteristics
Given their up close and personal nature, qualitative interviews naturally raise questions about differences
between interviewers and interviewees (Manderson, Bennett, & Andajani-Sutjaho, 2006). Some researchers
advocate matching by sex, age, race, and so on. Defenders of matching cite greater acceptability and
understandability because certain topics can be more easily addressed. In the Harlem Mammogram Study in
which I was a co-investigator, we recruited African American women as interviewers to maximize
participants’ comfort level when discussing sensitive issues surrounding bodily functions, sexuality, and racism
in health care.

Interviewer characteristics can make a difference. Miller and Glassner (2011) found different interview
responses from young men boasting about sexual experiences when the interviewer was male versus female, for
example, male interviewers did not question the men’s assertions that an incident of group sex with a woman
was consensual. When a young African American woman from their community interviewed the same men,
the latter were less boastful and more circumspect about the nature of the woman’s consent. Similar
discrepancies could occur when sexual orientation, age, gender, and social class distinguish the interviewer
from the interviewee. It is difficult to argue against older female interviewers in a study of women going
through menopause. A study of gay men as parents might be less productive if the interviewer is straight or
not a parent.

The decision to match often makes sense given the topic, but it is not always easy to make that call. And, once
matching is deemed important, which interviewer characteristics help or hinder a study? Robert Weiss (1994)
noted,

When I interviewed men who were IV drug users, I was an outsider to the drug culture but an
insider to the world of men. When I interviewed a woman who was an IV drug user living in a
shelter and also the mother of two children, I was an outsider to the world of women, drug users,
and women’s shelters, but an insider to the concerns of parents. (p. 137)

Using members of the study population and community as interviewers is laudable (and an essential ingredient
of participatory action research), but it carries risks as well. Respondents may fear a loss of privacy by speaking
to “one of their own,” or they may slant their responses to avoid loss of face with a peer. In the NYRS, we
could not match by gender (the majority of study participants were male and our staff female), but this did not
prove to be a problem in getting cooperation and rich interviews. Prior training and experience with this
population, along with respect and mutual understanding of what the study entailed, helped overcome gender
and other differences. For some studies, the effectiveness of the interview may depend on matching, but for
most, being a skilled interviewer is sufficient.

189
190
Common Problems and Errors in Qualitative Interviewing
Ideally, the qualitative interviewer has a broad fund of knowledge that makes probes productive and the
maturity to be patient, to know when to make these departures and when to remain silent. Yet even veteran
interviewers can have that sinking experience of losing control, or “taking a puppy for a walk” (Steinmetz,
1991, p. 64). Similarly, reining in an unruly focus group can be like trying to herd cats. Online interviews can
be cut off by the interviewee or by a lost Internet connection (and one cannot easily discern which is the
reason). Online video conferencing can bring delayed responses, interference from weak Internet signals, and
poor audio quality.

Aside from technological and human frailties, some of the common pitfalls of qualitative interviewing can be
anticipated. One occurs when the desire to control or lead compels the interviewer to interrupt the narrative
flow. Few things are more disappointing than reading a transcript (or listening to a recording) in which the
interviewer dominates the proceedings and cuts off the interviewee.

What about occasions when the interviewee is uncooperative? It is frustrating to sit across from someone who
answers questions in monosyllables, then sits impassively waiting for the next question. The interview stops
and starts, frustrations rise, and the interviewer is at her wit’s end. Respondents in these situations can be
exasperating, but the best tactic is to remain calm, be diplomatic, and stop if necessary. No amount of
information is worth risking coercion (or a migraine headache). If the resistance appears transitory, another
interview can always be scheduled.

Finally, inadequate training and fatigue can lead to slippage and unproductive improvisation. Either through
carelessness or well-intentioned ad-libbing, poorly prepared interviewers can reproduce all of the pitfalls
previously mentioned. Interviewers sometimes get a little too zealous in probing, especially when skeptical
about a participant’s claim. They could also become too familiar, and the interview becomes conversational
and veers off the topic. Once, when perusing transcripts, I came across an interviewer who prefaced a question
about drug use with “Sorry, but I have to ask this next question.” This seemingly minor act of distancing
herself from the study sent a message of not taking the question seriously, thereby practically inviting
participants to fudge the truth.

During a successful interview, qualitative researchers strike a balance between the general and the particular,
the need to stay on task versus the need to probe deeper. Simultaneously, the interviewer is expected to listen
empathically, monitor body language, anticipate the next question, and mentally or literally take note of red
flags (e.g., discrepancies, statements signaling deeper meaning). When everything is clicking, both interviewer
and interviewee part company feeling they have had a mutually beneficial encounter. Even a less-than-perfect
qualitative interview leaves most interviewees gratified by the experience of being respected and listened to.
And it often leaves interviewers gratified but drained emotionally.

191
192
Emotional Issues in Interviewing
The sensitive nature of qualitative research almost guarantees that emotionally laden information will surface.
Study participants may laugh, cry, or grow angry during an interview. Levy and Hollan (2000) write about
exaggerating the fragility of the interviewee as a recurring problem for novice researchers. Sensitive topics and
traumatic stories can bring painful emotions out; interviewers should never gratuitously probe or show
insensitivity in other ways. Yet to avoid addressing this information simply because it is emotion-laden
deprives the study and assumes that participants are incapable of handling their emotions. Indeed, the vast
majority welcome the opportunity to tell their stories to an empathic, nonjudgmental listener.

As mentioned in Chapter 4 on research ethics, emotional displays can happen, but they rarely cause more than
momentary interruptions. Even when interviews prompt strong reactions, a skilled interviewer can show
concern and then gently steer respondents to a calmer state (Weiss, 1994). In any event, the qualitative
interviewer does not try to elicit strong emotions, only to create a safe space for their expression if they need to
occur.

The interviewer can also experience a backwash of emotions following an intensely personal and painful
encounter. This can be exacerbated when the participants or the setting are familiar or the researcher has
shared experiences with his participants. A researcher interested in elder abuse after witnessing maltreatment
of his grandmother is not immune to reliving that pain during and after interviews. Unlike the relationship
between practitioner and client, the qualitative interviewer does not have the protection of the clinical remove,
settling instead for an “emotional middle distance” (Weiss, 1994, p. 123). One important way to maintain this
is through debriefing (more about this in Chapter 8). In the NYRS, the team met weekly to debrief about
each interview that occurred in the previous week. Although most interviews take place without incident, the
infrequent exception is worth discussing further, whether it is a study participant or interviewer who needs
help.

193
Interviewing in a Non-English Language
As discussed in Chapter 3, the explosive growth in qualitative methods in the United States and other
English-speaking countries has diverted attention from the salience of language differences. As methods of
qualitative analysis became more dependent on deriving meaning from texts, venturing beyond the English
language complicated research designs and sampling (Esposito, 2001; Shibusawa & Lukens, 2004; Twinn,
1997).

Although some qualitative methods do not work well with linguistic distance (e.g., phenomenological
interviewing), most can be used with intervening translation. Care must be taken, however, to reduce
misunderstandings and mistranslation as much as possible. A graduate student related to me her experience
interviewing women in remote villages in India where only tribal languages were spoken, thus requiring a
chain of translators (English to Hindi to tribal language and back). Every step increases the risk of distortion,
but there is no realistic alternative, at least in the short term. The optimal resolution involves recruiting and
training native speakers to collaborate as partners in the research rather than language technicians.

Noting the overwhelming influence of the English language on qualitative methods and studies should not
lead one to overlook the fact that non-English-speaking qualitative researchers have pursued studies with
noteworthy success, and their numbers are increasing each year. One need only peruse the International
Journal of Qualitative Methods (http://www.ualberta.ca/~ijqm) to see articles published in Spanish, Portuguese,
French, Russian, and Tagalog.

194
Interviewing Across the Digital Landscape
Notwithstanding the cherished status of in-person encounters, there are occasions when these are not feasible
or not the point. E-mail and other forms of electronic interactions—often mediated by a smartphone or tablet
electronic device—have risen in popularity in qualitative research (Hamilton & Bowers, 2006; Seale,
Charteris-Black, McFarlane & McPherson, 2010). There are, of course, differences between modes of
communication. Telephone and Internet video interviews share with in-person interviews the need to
schedule the encounter, but they have the advantage of being audiotaped, thus retaining access to voice and
intonation (video adds visual content). Telephone and Internet interviewing are a good fit for specific types of
studies, for second (or later follow-up) interviews and for occasions when respondents live far away or are
unavailable for other reasons.

Though still heavily dominated by surveys where the convenience is hard to beat, Internet-based research has
also become the province of qualitative inquiry. The Internet offers new ways to reach and communicate with
isolated and stigmatized populations as well as with technologically savvy respondents who prefer the
anonymity of being online rather than in person (Illingworth, 2001; Markham, 2005). Unlike in-person
interviews, Internet communications do not have to be scheduled (although this is an option) and are more
convenient to respondents sitting in the comfort of their home, office, or local cafe (Miller & Slater, 2000).
Without the pressure of time, users can compose, redraft, and think through what they want to say (Kosinki,
Matz, Gosling, Popov, & Stillwell, 2015).

Internet forums and chat rooms are free flowing conversations where users share everything from mundane
routines to intensely personal revelations. Their length has become ever smaller due to the popularity of
Reddit, Twitter, Facebook, Instagram, and other social media that curtail space in favor of sound or visual
bites. New ethical questions arise with this format (Flicker, Haans, & Skinner, 2004): Are these posts
considered private communications? Is it ethical for a researcher to join a chat room or conversation thread if
this is the only way to gain access? What if the researcher is already a member of the online group? Internet
postings are in the public domain, but users intend to share them with “desired others,” so their availability for
use as data is hardly a foregone conclusion. Given the growth in Internet research over the past decade, most
human subjects committees have become adept at answering such questions, balancing a presumption of
anonymity with concerns for privacy.

For example, men who have sex with men (MSM) are a favored focus of Internet research as these individuals
frequently use online contacts for sexual hook-ups and to engage in chats where anonymity permits frank
discussion and revealing photos (Ayling & Mewse, 2009). If the researcher adopts a passive, non-interactive
stance, these forums could become a vast trove of documents for review. (It is not acceptable ethically for the
researcher to use deception to enter these forums, but passive deception is a gray area.) As shown in Box 5.8,
some online forums facilitate community building and peer support.

One small caveat is that although the situation is changing by the day, Internet and smartphone access are not
universal and could exclude certain low-income groups. Cell phones are ubiquitous in low-income

195
neighborhoods, but Internet and WiFi connectivity are more costly. Homeless youths are adept at accessing
public library computers for Internet access (Rice & Karnik, 2012) and sharing inexpensive cellphones (Rice,
Lee, & Taitt, 2011). As with other mediums of communication, researchers seek to fit the chosen venues and
populations to the study’s goals.

196
Box 5.8 Internet Research on Persons with Early-Onset Alzheimer’s
Disease
The benefits of the Internet for homebound or disabled individuals include access to information and support available 24/7 from
Internet users residing all over the world. Rodriquez (2013) studied the empowering aspects of one such online community—adults
under age 65 with early-onset Alzheimer’s Disease (EOAD). Interestingly, his university human subjects committee did not deem
the study to involve “human subjects,” but Rodriguez nevertheless changed the screen-user names (self-chosen pseudonyms),
substituting his own pseudonyms. The study’s overall goal was to explore how persons with EOAD create a new “self” after the
diagnosis. Often losing their ability to drive or work before their mental faculties were compromised, these individuals shared stories
of their diagnoses, symptoms, family reactions, and hopes and fears about the future. Rodriquez studied 354 posts by 32 forum
members, coding and analyzing thematic content on topics such as “telling others” and “defining moments.” Examples of the latter
included giving up one’s driver’s license and sharing the diagnosis with coworkers. Part of a burgeoning literature on illness
narratives, Rodriquez’s report portrayed individual stories set amidst a supportive online community.

197
Use of Audio and Video Recorders
A preference for audio recording over note taking in interviews is unquestionable—the latter should be
reserved for occasions when recording is not possible. Audio recording captures what is said verbatim, but it
also captures ambient sound—children crying, ambulance sirens, and phones ringing—as well as vocal
intonations (laughter, crying, sarcastic tone). Although one might reasonably assume that some people do not
want to be recorded, this has proven to be a rare occurrence in recent years.

The superior technology and lower costs of digital voice recorders and smart phones have made audio
recording accessible for most budgets. Audio recorders typically come with the software to enable one to
download the interview onto a computer and transcribe it directly using controls displayed on the screen (with
headphones plugged into the computer). Although longed for by weary transcriptionists and resource-tight
qualitative researchers, voice recognition software is not up to the task of faithfully recording the nuances of
participants’ speech as of this writing.

Video recording of interviews is far less common, in part because of the risk of loss of privacy and
confidentiality. The advantages of video recording are noteworthy if the study participants (and one’s
institutional review board) are amenable. While the future may bring the same acclimation to video as to
audio recording, current practice reserves video recording for studies where observation is important in itself.
Two hypothetical examples come to mind: a study of children and playground bullying or a discourse analysis
focusing on adolescent-parent communication styles.

198
Using Documents and Other Archival Materials
A variety of documents and materials are of interest to qualitative researchers (Bowen, 2009). These include
printed matter such as court records, case reports, minutes of meetings, brochures, diaries, photographs,
letters, and so forth. Archived videos, films, and photographs are also useful. Documents and existing
information have distinct advantages including their lack of reactivity. Use of documents is also less time
consuming and emotionally taxing compared to the hands-on labor of observation and interviewing.

Some types of documents and archival materials have disadvantages because they were not produced for
research purposes. They may be inaccurate, uneven, or incomplete. This may be due to sloppy record keeping,
but it can also be deliberate. For example, minutes of staff meetings may be doctored to cover up embarrassing
revelations, or a physician may omit mention of a psychiatric history in a medical file to protect her patient
from stigma. Clearly, a study dependent on accurate documents is constrained by what is available and its
quality (Wolff, 2004). On the other hand, the naturalness and unfiltered quality of documents such as diaries,
online chats, and personal artwork can be what is valued most. A constructionist or interpretivist approach
would render such naturalness an essential quality.

Access to existing documents varies. Individuals may freely share personal documents, but privacy protections,
copyright laws, and other restrictions may apply to some materials. For example, the enactment of the Health
Insurance Portability and Accountability Act (HIPAA) in 1996 made medical records retrievable only with
consent from the patient. Although HIPAA restrictions are a hurdle to overcome, obtaining participants’
consent to access their medical records usually entails no more than adding a coda to the existing consent
form.

As mentioned earlier, Internet communications are a growing source of archival data. Blog, listserv, and chat
room conversations can be analyzed in the same way as hard copy documents, although ethical considerations
and proper approvals must be given priority because such information can be poorly protected and secured
(Hessler et al., 2003). In addition, the threshold between Internet interviewing and “existing data” can get
blurry. In their study of diaries kept by adolescents and sent via e-mail, Hessler and colleagues (2003) avoided
direct elicitation and attempted to be as unobtrusive as possible; the diary entries were aggregated and treated
as archival materials. In part because of uneven availability and depth, documents and archival materials tend
to be used as supplementary rather than primary in qualitative research.

199
Ending Data Collection
The decision to end data collection depends on a number of factors, methodological and practical. With
regard to the former, closure usually comes when saturation has been achieved (i.e., when the data show
redundancy and reveal no new information). This somewhat vague prescription is juxtaposed against
practicalities such as deadlines imposed by the study’s sponsors or resource constraints. Studies with relatively
homogeneous domains and/or sampling strategies are likely to reach saturation sooner than those with
broader reach and ambition. The ultimate goal is evidentiary adequacy based upon sufficient time in the field
and a corpus of data deep and varied enough to sustain the study’s analytic ambitions (Erickson, 1986;
Morrow & Smith, 1995).

Two aspects of data collection are peculiar to the timeline of qualitative research—the flexibility of the design
and the likelihood of return visits to the field (Iversen, 2008). Qualitative studies resist arbitrary endpoints.
There are the usual exigencies, for example, transcription delays, participants become elusive, or the researcher
gets distracted by other life responsibilities. Some qualitative methods, most notably ethnography and
community-based participatory research, need longer periods of time to come to fruition. In this context, the
relationship between researchers and participants may ebb and flow depending on the stage of the study and
the availability of resources to carry it out.

Ending data collection in qualitative research can be an emotional transition given the relationships that came
before. Study participants may regret losing the respectful camaraderie, and many a qualitative researcher has
felt bereft by study’s end (although we should not assume that participants need or want us to stick around).
Even here, the end may not be absolute. Study participants sometimes ask if they can call to say hello or to
check on the progress of the study—this is a reasonable request that can be honored.

Among the qualitative approaches, CBPR and ethnography instigate longer-term relationships that can
extend into years, so closure may be gradual and prolonged yet still difficult. Our experiences in the NYSS and
NYRS were bittersweet—relieved to be through with the emotional ups and downs of data collection but
sorry to lose the connections we had with participants. Most accepted the end with grace, and a few called in
the months after to say hello or share a life event.

200
Summary and Concluding Thoughts
This chapter has covered the expanding portfolio of data collection methods in qualitative research. The
immersion and intensity of qualitative inquiry make observation a uniquely valuable part of data collection.
When a qualitative study is focused on place and naturally occurring behavior, observation becomes crucial
(Tjora, 2006). There is bound to be an impact on what is being observed and what gets recorded, and the
researcher tries to be as unobtrusive as possible and use prolonged engagement to reduce reactivity.

Qualitative interviewing occurs in different ways and contexts, but its signature format is in-person,
individually or in small groups. Advance preparation involves developing and pilot testing an interview guide,
familiarizing one’s self with its contents, and practicing the gentle art of probes, both planned and
spontaneous. Conducting the interview is an exercise in multitasking, that is, asking questions, observing, and
unobtrusively taking notes on follow-up questions. Audiotaping relieves the interviewer of the onerous (and
error-ridden) task of note taking. When possible (and most of the time it is), each participant is interviewed
more than once to ensure depth and completeness.

The third type of qualitative data—documents and other materials—are the tangible and nonreactive
byproducts of human activity. Documents originate from a number of places, both personal and
organizational. Some archival materials raise understandable concerns about their accuracy and completeness.
For researchers more interested in interpretations based upon naturalness and authenticity, documents—
especially personal diaries, artwork and the like—are invaluable resources.

Data collection ideally comes to an end when saturation or redundancy has been achieved, but external factors
such as deadlines and diminishing resources can limit the time allotted for data collection. The emotional
sequelae of qualitative data collection reverberate throughout the study, often coming into sharp relief as it
draws to an end. Simply paying attention to the possibility (or inevitability) of such feelings and addressing
them with sensitivity can make closure much easier for all concerned.

201
Exercises

202
Observation
1. An Hour in the Life . . .

Spend an hour or more in the “field” as an ethnographer. This may be in a park, playground, cafe,
subway station, or sports arena—any public space where behavior can be unobtrusively observed. As you
carry out your observations, take in-depth field notes. Report on the physical space, actors, behaviors,
emotions expressed, and ambience of the setting (time of day, unusual events, etc.). After your field
notes are written, bring them to class, and share with a fellow student. Working together, place brackets
around portions that appear more interpretive than descriptive. In other words, try to separate
straightforward reporting from assumptions or biases.
2. Memory Fails . . . Field Note-Taking

The class instructor can show 10 to 15 minutes of a movie or video to students, and ask them to write a
description of what they have seen 24 hours later. Meeting in groups at the next class, students compare
their “field notes” and talk about the selectiveness of memory. Show the movie excerpt again, and then
discuss the importance of taking field notes as soon as possible after the observation.

203
Conducting an Interview
This exercise exposes the student to the intensity and flexibility of a qualitative interview. It also vividly
illustrates how time-consuming interviewing and transcribing are.

1. Choose a topic of interest, and seek out a knowledgeable respondent for an open-ended interview.
Develop an interview guide of four to five questions.
2. Conduct the interview using an audio recorder or smartphone. Make sure it lasts at least 30 minutes.
3. Transcribe the interview verbatim.
4. Look through the transcript critically, and note areas needing improvement. For example, did you
interrupt the flow? Ask leading questions? Become too conversational?

204
Using Available Documents
Think about your workplace or another familiar work setting as the object of an ethnographic study. Now
consider that you have been asked to study this place only using documents, archival materials, and so on—no
observation or interviewing allowed. Address the following:

1. Develop a list of the types of documents available.

2. How would you evaluate the relative quality and quantity of these materials?
3. Do you have any concerns about their accuracy and completeness? If so, what are they?

205
Additional Readings

206
Interviewing
Atkinson, R. (1998). The life story interview. Thousand Oaks, CA: Sage.

Ayling, R., & Mewse, A. J. (2009). Evaluating Internet interviews with gay men. Qualitative Health Research,
19, 566–576.

Baker, S.E., & Edwards, R. (2012). How many qualitative interviews is enough? London, UK: National Centre
for Research Methods. Retrieved from http://eprints.ncrm.ac.uk/2273/4/how_many_interviews.pdf

Best, S. J., & Krueger, B. S. (2004). Internet data collection. Thousand Oaks, CA: Sage.

Dick, H. P. (2006). What to do with “I don’t know”: Elicitation in ethnographic and survey interviews.
Qualitative Sociology, 29(1), 87–102.

Fontana, A., & Frey, J. H. (1994). Interviewing: The art of science. In N. K. Denzin & Y. S. Lincoln (Eds.),
Handbook of qualitative research (pp. 361–376). Thousand Oaks, CA: Sage.

Gubrium, J. F., & Holstein, J. A. (Eds.). (2002). Handbook of interview research. Thousand Oaks, CA: Sage.

Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data
saturation and variability. Field Methods, 18(1), 59–82.

Hamilton, R. J., & Bowers, B. J. (2006). Internet recruitment and e-mail interviews in qualitative studies.
Qualitative Health Research, 16, 821–826.

Hammersley, M., & Gomm, R. (2008). Assessing the radical critique of interviews. In M. Hammersley,
Questioning qualitative inquiry (pp. 89–100). London, UK: Sage.

Hewson, C., Yule, P., Laurent, D., & Vogel, C. (2003). Internet research methods: A practical guide for the social
and behavioral sciences. Thousand Oaks, CA: Sage.

Holstein, J. A., & Gubrium, J. F. (1995). The active interview. Thousand Oaks, CA: Sage.

Irwin, L. G., & Johnson, J. (2005). Interviewing young children: Explicating our practices and dilemmas.
Qualitative Health Research, 15(6), 821–831.

Krueger, R. A., & Casey, M. A. (2014). Focus groups: A practical guide for applied research (5th ed.). Thousand
Oaks, CA: Sage.

Kvale, S. (1996). InterViews: An introduction to qualitative research interviewing. Thousand Oaks, CA: Sage.

Manderson, L., Bennett, E., & Andajani-Sutjaho, S. (2006). The social dynamics of the interview: Age, class
and gender. Qualitative Health Research, 16(10), 1317–1334.

McCracken, G. (1988). The long interview. Newbury Park, CA: Sage.

207
Miller, J., & Glassner, B. (2011). The inside and the outside: Finding realities in interviews. In D. Silverman
(Ed.). Qualitative research: Issues of theory, method and practice (3rd ed.), pp. 291–309. London: Sage.

Morgan, D. L. (1997). Focus groups as qualitative research. Thousand Oaks, CA: Sage.

Richardson, L. (2002). Poetic representation of interviews. In J. F. Gubrium & J. A. Holstein (Eds.),

Handbook of interview research: context and method (pp. 78–89). Thousand Oaks, CA, Sage.

Rubin, H. J., & Rubin, I. S. (2011). Qualitative interviewing: The art of hearing data (3rd ed.). Thousand
Oaks, CA: Sage.

Salmon, A. (2007). Walking the talk: How participatory interview methods can democratize research.
Qualitative Health Research, 17(7), 982–994.

Seale, C., Charteris-Black, J., McFarlane, A., & McPherson, A. (2010). Interviews and Internet forums: A
comparison of two sources of qualitative data. Qualitative Health Research, 20, 595–606.

Seidman, I. (2006). Interviewing as qualitative research. New York, NY: Teachers College Press.

Spradley, J. P. (1979). The ethnographic interview. New York, NY: Holt, Rinehart & Winston.

Weiss, R. (1994). Learning from strangers: The art and method of qualitative interview studies. New York, NY:
Free Press.

Wimpenny P., & Gass, J. (2000). Interviewing in phenomenology and grounded theory: Is there a difference?
Journal of Advanced Nursing, 31(6), 1485–1492.

208
Observation (including Participant Observation)
DeWalt, K. M., & DeWalt, B. R. (2001). Participant observation: A guide for fieldworkers. Walnut Creek, CA:
AltaMira Press.

Emerson, R. (2001). Contemporary field research: Perspectives and formulations. Long Grove, IL: Waveland
Press.

Emerson, R. M., Fretz, R. I., & Shaw, L. L. (2011). Writing ethnographic field notes (2nd ed.). Chicago, IL:
University of Chicago Press.

Kusenbach, M. (2003). Street phenomenology: The go-along as ethnographic research tool. Ethnography, 4,
455–485.

LeCompte, M. D., & Schensul, J. J. (1999). Designing and conducting ethnographic research (Ethnographer’s
toolkit, Vol. 1). Walnut Creek, CA: AltaMira Press.

Lee, R. M. (2000). Unobtrusive measures in social research. Philadelphia, PA: Open University Press.

Lofland, J., & Snow, D. A. (2005). Analyzing social settings: A guide to qualitative observation and analysis.
Belmont, CA: Wadsworth.

209
Analysis of Documents and Archival Data
Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9, 27–
40.

Wolff, S. (2004). Analysis of documents and records. In U. Flick, E. von Kardoff, & I. Steinke (Eds.), A
companion to qualitative research (pp. 284–289). London, UK: Sage.

210
Visual Media (including Photo-Voice and Photo-Elicitation Interviewing)
Collier, J., & Collier, M. (1986). Visual anthropology: Photography as a research method. Albuquerque: University
of New Mexico Press.

Epstein, I., Stevens, B., McKeever, P., & Baruchel, S. (2008). Photo elicitation interview (PEI): Using photos
to elicit children’s perspectives. International Journal of Qualitative Methods, 5(3), 1–11.

Han, C. S., & Oliffe, J. L. (2015). Photovoice in mental illness research: A review and recommendations.
Health, 1–17.

Harper, D. (2002). Talking about pictures: A case for photo elicitation. Visual Studies, 17(1), 13–26.

Lapenta, F. (2011). Some theoretical and methodological views on photo-elicitation. In E. Margolies & L.
Pauwels (Eds.), The SAGE Handbook of Visual Research Methods, London (pp. 201–213). Thousand Oaks, CA:
SAGE.

Padgett, D. K, Smith, B. T., Derejko, K., Henwood, B. F., & Tiderington, E. (2013). A picture is worth . . .
? Using individual photo-elicitation to enhance interviews with vulnerable populations. Qualitative Health
Research, 23, 1435–1444.

Rose, G. (2007). Visual methodologies: An introduction to the interpretation of visual materials. London, UK:
SAGE.

Wang, C., & Burris, M. A. (1997). Photovoice: Concept, methodology, and use for participatory needs
assessment. Health Education & Behavior, 24(3), 369–387.

Wang, C. C., Yi, W. K., Tao, Z. W., & Carovano, K. (1998). Photovoice as a participatory health promotion
strategy. Health Promotion International, 13(1), 75–86.

211
6 Data Analysis

Qualitative data analyses are steeped in choices and decisions. They may hew closely to specific procedures or
venture into realms of imaginative artistry. All qualitative researchers have to contend with masses of raw data
that need to be reduced and transformed through an iterative process of reading, describing, and interpreting.
Pre-existing theories and sensitizing concepts may be brought to bear, but their survival in the final analysis
depends on the data. The balance between staying close to the data and thinking abstractly is a defining
feature of qualitative analysis.

An early decision is whether the researcher is seeking to explain something or develop a theory. This could be
done using analytic induction and grounded theory (GT) or one might use abductive thinking as described in
Chapter 1. Although GT has been criticized for yielding few actual theories (Timmermans & Tavory, 2012),
abduction remains relatively unknown and underused in qualitative methods. Researchers who eschew
explanation but still want to theorize have other routes to follow during analysis.

Qualitative methods employ an “arts and crafts” approach. However, their localized adaptability has come
with a price—an historic tendency to obscure the specifics of data analysis. While ethnography has retained
much of this aura of mystique, grounded theory opened the door to greater transparency in methods,
particularly in data analysis. Nevertheless, creativity and interpretation are still what make qualitative inquiry
the “art” as well as the “craft” that it is.

Analyses may range from fine-grained examinations of texts (via narrative and discourse analyses) to holistic
syntheses of multiple sources of data (case study and ethnography) to interview-intensive approaches falling
somewhere in between (phenomenological analysis and grounded theory). As a general rule, the less
structured the interview, the more likely the data analyses will be nonprescriptive. This chapter begins with
the logistics of management and preparation of data and then provides an overview of various analytic
approaches. Greatest emphasis will be placed on the type of analysis most common in qualitative methods—
coding and thematic development.

Interpretation—elevating the analyses into broader realms of meaning—is addressed in Chapter 7. As with
the iterative relationship between data collection and analysis, the boundaries between analysis and
interpretation are permeable and may be crossed back and forth as the study progresses toward completion.
Thus, it is somewhat artificial to divide these essential activities into two chapters, but their importance
deserves the additional space.

212
Data Management: Dealing With Volume Early On
Proper management is essential given the massive amount of raw data needing storage and accessibility for
retrieval. The tasks begin with fully disguising participants (usually with ID numbers) in all transcripts, audio
files, field notes, and other documents. When case study analyses are used, an ID number helps to keep the
various data sources linked together for each particular case. Inexperienced qualitative researchers (and most
quantitative researchers) are surprised by the sheer quantity of raw data generated by studying a small number
of people, that is, hundreds if not thousands of pages of transcripts, field notes, documents, and memos.

213
Using Qualitative Data Analysis (QDA) Software
QDA software has acquired a “must-have” status for many researchers, lending technological cachet to a
methodology known for being low tech and low threshold. The traditional tasks of manually cutting and
pasting have given way to dedicated QDA software such as ATLAS.ti, NVivo, and HyperRESEARCH.
Recent additions to the software possibilities are DeDoose, QUIRKOS, QDA Miner, and QDAP. These
QDA options may be purchased or used via monthly user fees; most offer discounted student versions as well
as free trials. Potential buyers should browse user feedback online as some are better suited to content analysis
and word or phrase searches than for more advance interpretive operations. (See further information at the
end of this chapter.)

The regularity of QDA software upgrades and the learning curve required to use them make it advisable to
offer a general overview and suggest additional readings and resources including free trial downloads. These
programs allow input of texts (transcripts, field notes, memos) as well as photos and video. The former are
typically entered in rich text format independent of Word or other programs. In ATLAS.ti for example, the
transcript file should be converted to pdf format if line numbers from Word documents are to be kept intact.
Compatibility features include exporting of files to or from SPSS or EXCEL.

The central functions of QDA software—to store data and facilitate coding and analysis—make it possible to
search for connections or hierarchies among codes, to produce graphical displays of codes, and to easily
retrieve information in an organized fashion. Coding on-screen is made easier by using a drop-down menu
displaying the codebook and by the software’s capacity to handle multiple overlapping coded sections. Auto-
coding (text searching for code words) and key word in context (KWIC) searches can easily be performed,
similar to using the “find” command. In the (New York Services Study) NYSS, for example, we used
ATLAS/ti to search for all versions of the word help (the wildcard designation allowing recognition of
multiple forms such as helped, helping, etc.). As we soon discovered, this type of searching yields hits as well
as misses, and the researcher must scrutinize the results to weed out the latter. The statement “my therapist
helped me overcome my fear” is meaningful; the phrase “I helped myself to another free t-shirt” is not. Since
autocoding and KWIC cannot substitute for the cognitive functions of inductive analysis, they are more
appropriate for content analysis (and even here must be checked for errors). Boolean search functions allow a
more targeted search for words or phrases using the and, or and not qualifiers.

The decision about whether to use QDA software usually rests on the scope of the project (including its
budget) and the researcher’s comfort level. Even a skeptic about software can find it useful when juggling so
much data and ideas simultaneously (Staller, 2002). A study with substantial amounts of data and multiple
users makes the cost of the software worthwhile, and tech-savvy users prefer the ease such software affords.
Student rates are affordable and worth the investment if the plan is to use the software over time. Meanwhile,
those less technologically inclined can still accomplish the tasks of qualitative data analysis with computer
office programs such as Microsoft Word.

Figure 6.1 shows a screenshot view of a New York Recovery Study (NYRS) transcript in ATLAS.ti with

214
rudimentary codes attached in the right-hand column. These code labels—social relationships, housing,
services, and substance use—were deliberately kept broad for more refined coding and analysis later on. As
shown, ATLAS.ti has an icon-driven interface including pull-down menus for code labels (the codebook) and
memos.

Figure 6.1 Screenshot From ATLAS.ti: Excerpt From Transcript in the NYRS

215
Teams, Multiple Users, and Cross-Site Databases
Team-based qualitative studies have become more common as the amount of effort can be considerable, and
the multidisciplinary benefits are many (Guest & MacQueen, 2008). Having more than one researcher
independently coding the same data and then comparing results is a useful means of minimizing bias (more on
co-coding later in this chapter). When a team is involved, role definitions need to be spelled out and
complementarity assured, but the burdens are more easily shouldered this way (of course the longstanding
tradition of the lone researcher, especially the dissertation researcher, has not disappeared). Beyond larger
sample sizes, the expanded scope of a qualitative study might include multiple sites or periods of data
collection.

Large databases with multiple users have long been the province of quantitative studies, and their
standardization facilitates tasks such as entering data from different sites, managing and cleaning the data,
merging files, and making subsets of data available to multiple users. Large-scale qualitative (and mixed
methods) studies bring more challenges but also reap tangible benefits (Manderson, Kelaher, & Woelz-
Stirling, 2001).

The site specificity and flexible designs of qualitative research can hinder large-scale collaboration, but it is
possible with proper planning and sufficient resources. At a minimum, advance thought needs to be given to
coordination across sites to ensure that datasets can be integrated. QDA software can be invaluable for these
transactions. The timing and scope of the cross-site collaboration lead to several questions: Will all data be
sent to a central repository or remain local? If the former, when will the data be sent and in what form? If the
latter, how will cross-site quality control be monitored? Will full collaboration be activated from the earliest
stage of the study or after the data have been collected? How will the collaborative team divvy up tasks?
Whereas data collection is inherently site specific, data analyses and writing may be integrated across sites
and/or take place at a single headquarters site. Finally, how will the allocation of responsibilities be reflected in
authorship of reports and publications? As a general rule of thumb, the more that localized interpretation is
valued, the less centralized will be the coordination. If the researchers want to (or must) produce cross-site
comparisons, some standardization of data collection, management, and analysis becomes necessary.

216
Transcribing Interviews
Transcription is an essential form of data filtering (MacClean, Meyer, & Estable, 2004). It is important to
train and supervise the study’s transcribers, including making them aware of the need for confidentiality.
Although not always possible, interviewer self-transcription is optimal and has several advantages, including
the ability to (1) fill in unclear passages; (2) insert explanations or clarifications; and (3) obtain timely feedback
on one’s interviewing technique.

There is no substitute for hearing one’s own voice and reliving the interview (especially with notes jotted
down to supplement the audio recording). For example, an interviewee might use gestures for emphasis or to
replace words altogether (e.g., a wink or a smile to mean “I was only joking,” a shrug instead of “I don’t
know,” or eye-rolling to signal impatience). Participants often tell animated stories in which they act out
scenes for the interviewer’s benefit. “Outside” transcribers are left scratching their heads in such instances. In
my experience, professional transcription services are unaccustomed to the nuance and attention that most
qualitative interviews bring, and many study participants (SPs) do not speak with crisp unaccented elocution.
Unfamiliar colloquialisms and pronunciation, in combination with slurring and soft speech, slow down the
process and cause errors. We have frequently had to re-listen to the tapes to fill in unclear passages or to
correct errors left by professional transcribers.

Be sure to develop basic rules for transcription, and ensure that they are followed consistently. These include
transcribing nonverbal utterances such as sighs, sobs, and laughter (setting these off with parentheses is
helpful). Pauses by the respondent lasting more than a few seconds are worth noting parenthetically or with
ellipses (. . .). Sometimes the interviewer needs to add a clarifying phrase in brackets so that future readers of
the transcript will not be confused or misled. In one of our earlier studies, one SP spoke repeatedly of “Susie,”
which led us to think he had a girlfriend, until the interviewer correctly identified “Susie” as SP’s dog in the
transcript. Sometimes an interviewee slurs incomprehensibly or talks very softly—this can be noted in brackets
as [unclear]. Another use of bracketing is to provide translation of foreign-language words or idioms.

The transcriber should studiously avoid editing and cleaning up grammar or off-color language. Respondents
have a right to have their stories transcribed without cosmetic and potentially distorting revisions. This
concern with fidelity is not the same as “triaging” transcription, that is, selectively transcribing to omit
tangential portions of the interview. Given time and labor costs, researchers may instruct transcribers to
overlook small talk or long-winded repetition. This, of course, is a decision that should be made with caution.

The risk of breaching privacy in transcription can happen when interviewees refer to other people and places
by name. Full names should not appear in the transcript but a first name or initials might suffice along with an
explanation in brackets, for example, JM [SP’s cousin]. At the same time, street addresses or place names
(e.g., Riverview Hospital) can be important to know and should be retained in the transcript. Although it is
essential to fully disguise all names in public presentations of the data and findings, it is better to retain such
details in the transcripts. Here are a few logistical suggestions for transcription:

217
 1. Leave ample margins if you plan to code or put notes on the hard copy of the transcript.
2. Number the lines sequentially from start to finish (some software programs do this automatically).
3. Use a small-font header on every page noting the interviewer’s initials, date of the interview, and date(s)
of the transcription.
4. Put interviewer questions in bold font to make it easier to scrutinize the content of questions as well as
their length.
5. Start every answer with the participant’s identification number so that any chunks of narrative moved to
new files will be identifiable.
6. It is okay to skip over the many “uh-uhs” and “umms.” (The exception being conversation analyses
where such utterances are important.) In the NYSS, participants said “you know what I’m sayin’” so
often that the transcribers typed YKWIS for shorthand.
7. Back up all work early and often, and keep back-up files in different places including cloud storage.

Given its intense labor, qualitative researchers frequently opt to pay others to do the transcription. This can be
expensive, but it also saves a lot of time. Transcribing a 90-minute interview can take 8 to 10 hours and
produce more than 30 pages. Using an outside transcriber brings the risk of errors and misunderstandings.
One compromise is for interviewers to transcribe their own interviews in the early phases and check the
outside transcriptions (chosen at random if not all transcripts can be checked). This is especially helpful in
filling in the unclear passages that the interviewer can more easily recall when listening to the audio recording.

In addition to encountering unfamiliar terminology, transcriptionists may fill in the blanks when the speech is
muffled or background noise intrudes. They may decide to edit out foul language or “mispronunciations.”
Sometimes it is difficult to know who is correct—the transcriber or the interviewee. (See Box 6.1 for a few
examples.)

218
 Box 6.1 A Brief Quiz on Transcription “Errors”
Interviewees may use colorful phrasing, and transcribers may err in capturing them or (worse) deliberately change the language. The
following quotes contain some examples of transcriber error interspersed with actual statements made by study participants in our
research. Can you detect which is which? (The Answer Key follows.)

1. “I was diagnosed as schizo-defective.”

2. “The doctor kept saying ‘take a seat,’ ‘take a seat,’ and I didn’t want to hear ‘take a seat.’”
3. “Someday I want to stay at the Plasma Hotel.”
4. “I am very bi-popular.”
5. “My roommate and I brush our teeth together.”

Answer Key: #1 and #3 were actual statements made by participants; #2, #4, and #5 were transcriber errors: #2 should read “Hep C”
(hepatitis C) instead of “take a seat,” #4 should read “I am very bipolar,” and #5 should read “my roommate and I watch TV
together.”

Interviews are entertaining and informative, but they can also be intensely personal—listening to teary
accounts or angry harangues can take its toll on the transcriber-as-listener. Like other members of the team,
transcribers can benefit from periodic debriefing to express their concerns and suggestions.

219
Translating and Transcribing in a Non-English Language
Aiming for accurate transcription becomes ever more complicated when the language of the participants does
not match that of the researcher. Translation into English (assuming this is the language of the researcher and
the ultimate report) can be done during the interview, later during transcription, or even later during analysis
and write-up. Each of these options brings risks of distortion, and the first can be disruptive to the flow of an
interview. There are few guidelines to follow here, but the usual approach is to translate as closely as possible
to the intended meaning—when words or idiomatic phrases do not translate, leave them intact (put in italics),
and explain their meaning in brackets or footnotes.

Errors in translation can result from a lack of familiarity with local dialects and meanings as well as deliberate,
even if well-intentioned, bias. Translators may feel that they need to safeguard their community’s reputation
by soft-pedaling negative statements. In addition to culturally idiomatic phrases, meaning arises from the
texture of speech—the words, cadences, and inflections that non-native speakers often fail to understand. For
example, the Japanese term amae has no English counterpart in its connotation of interdependency and
indulgence on the part of siblings caring for an ailing brother or sister (Shibusawa & Lukens, 2004). Concerns
about literal translation are further compounded by the inevitable nonverbal cues sent by facial expressions,
body language, and so on. As noted by anthropologist Clifford Geertz (1973), a wink may have many cultural
and situational meanings, or it may just be an involuntary twitch.

Both transcribers and translators should be part of the research team to the extent possible. They too are privy
to the intense human emotions evoked by qualitative interviewing and may benefit from debriefing. In
addition, sharing a deeper understanding of what the study is about helps translators and transcribers to
reduce errors and makes them feel valued as having a substantive contribution to make.

220
Qualitative Data Analysis: Beginning the Search for Meaning
Some qualitative researchers assert that findings are discovered as if they are lying in wait, and others say that
findings are social constructions. Beneath this semantic and epistemological divide is a common substrate of
activities that most often involves pattern recognition and thematic development. Such activities are
influenced by whether the study is concerned with change over time and whether its “cases” (individuals or
other units of analysis) are treated holistically or as part of an aggregate whose words or utterances constitute
the raw material of analysis.

Although all data are filtered in some way, qualitative data can be viewed as existing on a continuum based on
the degree of abstraction and processing involved. The least processed forms of data include audio files, visual
media (such as participant-produced photographs), and existing documents; partially processed data refer to
field notes, transcripts, and translations. For many qualitative studies, the next level of processing involves
coding—concepts or meaning units drawn from raw and partially processed data— followed by the
development of themes or categories. Parallel to data analysis and transformation are auditing and operational
tasks including memo writing of analytic decisions and journal keeping to record the researcher’s personal
reactions, biases, and concerns.

There is no substitute for diving into the corpus of data, reading and rereading transcripts, notes, and
documents. Inhabiting the data in this way ensures that the analyses are rooted in deep understanding. Memo
writing has moved to the forefront of data analysis as the connective tissue binding disparate activities and
helping to coalesce decisions that propel the analyses forward to conclusions. Memos inscribe thoughts,
reactions, and linkages. They help to crystallize inchoate hunches about what is going on in the data. Keeping
track of these memos, for example, logging them by date and time, contributes to an audit trail (more about
this in Chapter 8).

Boyatzis (1998) distinguishes between manifest and latent analysis, the former referring to surface description
and the latter to an interpretation of underlying or hidden meanings. One does not usually plunge into
analyzing the latent before gaining a comprehensive understanding of the manifest. Ethnographers, for
example, ensure they have ample description in order to make their interpretations “deep” and “thick” enough
to uncover the tacit meanings of cultural beliefs and practices.

At the same time, not venturing beyond the manifest is an error of omission, albeit one usually made with the
best of intentions (some researchers assume that looking into deeper meanings dishonors participants). In
writing about his study of homeless booksellers in New York City, Duneier (1999) notes, “if I had simply
taken the men’s accounts at face value, I would have concluded that their lives and problems were wholly of
their own making” (p. 343), thus overlooking the influence of larger political and economic forces. By the
same token, rushing to interpretation can render the data beside the point rather than the starting point.

Is there a happy medium? To paraphrase a noted social scientist, “you can’t generalize from the local, but you
can’t generalize without it” (Kotkin, 2002, p. B11). Readers of qualitative reports are asked to trust the

221
researcher to be a guide leading them into new understanding, and the researcher has to earn this trust. To do
so, we provide sufficient detail to assure readers of our deep familiarity with the setting and data, but we are
also obliged to make connections “up and out” to the larger context of opportunities and constraints. The
reader may be asked to take a “leap of faith” (Duneier, 1999, p. 343) in this regard, but the leap should have a
credible landing (much more about this in Chapter 7).

Whereas some methods offer specific albeit flexible guidelines for data analysis, others are less explicit in the
how-to aspects. Still other qualitative approaches, such as case studies and ethnography, exist as “meta-
methods” (i.e., broad enough to incorporate differing modes of data collection and analysis). Regardless of
approach, qualitative analyses depend on close and careful readings of texts, multitasking to attend to what
and how something is said or done, and using filters and analytic axes to organize the process as it unfolds.
Qualitative data analysis rarely follows a predictable course, so keeping track of its progress is critical.

222
Case Summaries
With their small sample sizes and intense involvement, virtually all qualitative studies can benefit from
making case summaries of each participant (and for case study analyses this is essential). Compiling case
summaries involves assembling and summarizing all available data about each particular case so that it may be
viewed holistically. SPs volunteer a lot of information—some important and some not; the case summary
organizes what is important for greater retrievability. In the NYRS, for example, case summaries could be
readily consulted to find out how many children a participant had, her place of birth, current substances being
abused (if any), and so forth. In addition to the routine demographic characteristics elicited at the end of the
interview, respondents volunteer factual information during interviews that can be included in the case
summary. I prefer to do this during the first cycle of coding, flagging this information in the left-hand column
of the transcript to be later entered into the case summary.

223
Theories and Concepts in Qualitative Data Analysis
Theorizing rather than theory testing reflects the dynamic way that researchers interact with ideas and
interpretations during analysis. Theorizing allows for conceptualizing without necessarily producing a theory
—the reality for the vast majority of qualitative studies. It also implies a dynamic process with varying inputs
and outputs depending on the study’s aims, research design, and stage of analysis.

The most prepackaged stance toward concepts and theories conforms to what Crabtree and Miller (1999) call
a template approach. Using this approach, a researcher relies on a codebook largely or entirely developed in
advance. Content analysis is closely associated with this option, but any qualitative study that needs to follow a
prescribed conceptual framework may go this route. At the other end of the continuum are studies that reject
the use of pre-existing concepts in favor of naive immersion in the data. A phenomenological approach, for
example, places high priority on exploring the lived experience de novo to reduce filtering and distortion that
can undermine authenticity.

Along the continuum’s middle ground lie qualitative studies such as grounded theory. Thus, the researcher
may use sensitizing concepts from existing theories, but their place in the findings is by no means guaranteed.
In this way, the sine qua non of qualitative research—its capacity for surprise and new insights—remains
intact. For the most dedicated grounded theorists, the development of a theory or model (whether having low,
medium, or high explanatory power) is the desirable end result.

224
Data Analysis in Diverse Qualitative Approaches
All data analyses eventually circle back to the research questions that set the stage for the study in the first
place. However, answering these questions is only part of the process since qualitative studies have the
capacity to deliver much more. More includes new information and fresh perspectives as well as noting what
was not seen or not said. It is important to reflect back on one’s epistemological stance and specific approach
to enhance the coherence of the study’s analyses.

One of the most common activities across qualitative approaches is the pursuit of themes, and here confusion
often reigns. The desire to develop themes is understandable, especially when one goes beyond scrutinizing a
single case or portion of narrative. The attraction of thematic analysis is readily apparent, as it is accessible and
exists independently of theoretical or epistemological frameworks (Braun & Clarke, 2006). As such, thematic
analysis has an all-purpose quality that makes its adoption or uptake by more formalized approaches relatively
easy. GT, phenomenology, case studies, and narrative analyses often use thematic analysis to develop theory,
identify the essences of a lived experience, and address commonalities across cases or narratives. In the
following sections are general descriptions of data analysis used in the different qualitative approaches.

225
Content Analysis
Content analysis has a distinct history in communications, journalism, and business (Berelson, 1952). It was
originally developed to count the number of incidents of the phenomenon of interest. Content analyses of
newspapers, magazines, television, and Internet communications could be used to reveal, for example, the
frequency of pharmaceutical advertisements, the number of violent incidents involving children, or the
prevalence of racial stereotyping. Content analysis is not necessarily atheoretical, but its application is largely
one of documentation rather than interpretation.

In recent years, content analysis has evolved toward a hybrid approach designed to “distill words into fewer
content-related categories” (Elo & Kyngäs, 2008, p. 108). Working in the margins between quantitative and
qualitative methods has brought criticism to content analysis—statisticians disparage its simplistic analytic
techniques, and qualitative researchers note its reductionism as a lost opportunity (Morgan, 1993). In keeping
with this eclecticism, content analysis can be applied to quantitative or qualitative data and can be used
deductively as well as inductively (Elo & Kyngäs, 2008; Hsieh & Shannon, 2005; Mayring, 2004). In the first
stage known as preparation, the unit of analysis—a word, phrase, or sentence—is identified. If the content
analysis is inductive, the analyses take a familiar step toward coding and categorization. If taking a deductive
approach, the researcher arrays the pre-existing categories in a matrix and fills it out with content that fits
(Elo & Kyngäs, 2008). Data of interest to the study that do not fit into the matrix can be used to create new
categories (switching to inductive mode). This alternating between inductive and deductive is common to
content analysis (Sandstrom, Willman, Svensson, & Borglin, 2015).

The reader might ask what is different about current practices in content analysis compared to the other
qualitative approaches, and this author is not entirely sure of the answer. In the world of qualitative methods,
many approaches coexist and overlap, yet somehow maintain their integrity and distinctiveness. To the relief
of some and vexation of others, qualitative methods place a premium on being diverse and boundary
transgressing.

226
Ethnography and Data Analysis
Regardless of whether they are conducted in Samoa or Springfield, ethnographic methods produce a wealth of
data that can quickly overwhelm the researcher unaccustomed to multitasking with minimal guidance. The
following quote describes this laborious tradition:

Working with these materials was a messy, exasperating, and complicated procedure. I began by
reading all the field notes and raw materials repeatedly until I knew what was in each volume and
where it was, creating a sort of mental map and table of contents. Then, as the structure and order
of presentation of topics became clearer, I literally surrounded myself with data. I made concentric
circles of important pages of field notes, articles, books, and drafts, and I perched in the middle of
these to think, sort, and combine. Each of these circles became a chapter, but only after it had
become a shambles. Days were spent shuffling and grabbing, realizing a whole section needed
rewriting and so beginning again, or rescuing all from numerous disasters with the paws of muddy
dogs who assaulted me for attention. (Estroff, 1981, pp. 33–34)

In addition to field notes, ethnographic data include interviews, documents, and records. They may also
include quantitative measures and analyses. As a rule, field notes are analyzed the same as interview
transcripts; analytic memos keep track of which type of data contributed what information. Estroff’s analyses
were the old-fashioned manual kind involving visually displaying data and intense contemplation about what
was being observed and interpreted. Whether done on the dining room table or virtually with a computer, the
principles are the same.

To get started analyzing ethnographic data, the researcher uses field note summaries, individual life stories,
and abundant memos to help structure the sifting and sorting process. He may triangulate different sources of
data (not to confirm but to build or expand meaning). Here is a hypothetical example of how multiple sources
of data can expand meaning. Consider a study set in a small-town Texas community where many Iraqi war
veterans and their families live. Although many are known to suffer from psychological trauma, few veterans
make use of a local mental health clinic. The literature points to stigma and fear of being labeled “crazy” as a
primary cause of this phenomenon. Ethnographic research at the site also reveals that the physical layout of
the clinic—long dim hallways and windowless offices—made already troubled veterans reluctant to use the
facility. Post-traumatic stress arising from chaotic war experiences can lead to feeling trapped with no exit,
and the layout of the clinic made them uneasy. Finally, in-depth private interviews with the veterans revealed
yet another reason: They feared a psychiatric diagnosis would lead to confiscation of their guns. Local culture
valorizes use of guns for hunting and target practice, and such a loss—however unlikely—would be deeply
resented. Taken together, these sources of data paint a more multifaceted understanding of why veterans
might avoid mental health treatment.

Ethnographic data are analyzed for idiographic detail, but they should ultimately address larger concerns.
These may be theoretical, critical, policy-relevant, or so on. Learning how to “sweep back and forth” and

227
“swoop in and out” of the data is one of the most challenging aspects of ethnographic data analysis. A number
of ethnographic experts have stepped forward over the years to instruct novices (see the reading list at the end
of this chapter), but all agree that strictly following a formula is not the way to go. In the end, the best way to
demystify ethnographic data analysis is hands-on experience along with documentation. Although this is true
of all qualitative methods, the learning curve in ethnography can be steep given its demands of prolonged
engagement in what is usually an unfamiliar setting.

228
Case Study Analysis
Like ethnography, case study analysis is a meta-method embracing multiple forms of data and analytic
techniques and lacking in codified procedures. As mentioned in Chapter 2, Stake (1995) distinguishes
between instrumental, intrinsic, and multiple case studies. Whereas instrumental case studies are illustrative
devices used to highlight discussion of a larger issue or concern, intrinsic case studies focus on the case itself as
worthy of intensive scrutiny. Multiple case study analysis follows the same principles of a single case study but,
for reasons of interpretation and expansion, extracts meaning from more than one case.

A primary feature of case study analysis is going “deep” before going “out” (to larger issues and theories) or,
for some studies, going “across” to other cases. Doing within-case analysis means delving into historical
background and/or exploring the case in all of its complexity. If very little is known about the case and it has
intrinsic interest, analysis may focus more on description than interpretation.

Because cases may be persons, entities, or events, analysis plans vary depending on the case and the data to be
collected. They also employ differing approaches depending on whether the study design is cross-sectional or
chronological. Patton (2002) notes that the choice of a case may shift during sampling (e.g., from an entire
school to selected classrooms or from classrooms to selected teachers and students). A case study may
document the inner workings of the case, and it may also advance an argument. An example of this is
Bradshaw’s (1999) study demonstrating that the closure of a military base (his case) did not produce the
predicted dire consequences. A case study may also explore the causes and consequences of a major policy
change, for example, the New York City Police Department’s shift to a “broken windows” policy of quality-
of-life arrests in the 1990s (Kelling & Coles, 1996).

In the NYRS, we assembled data for each participant (or case) covering the 18 months they were enrolled in
the study. This included four interviews (at baseline, 6, 12, and 18 months), individual case summaries, and—
for a subsample—shadowing interviews as well as photo-elicitation interviews. Data included transcripts, case
summaries, photographs, and field notes. A matrix was developed to display salient domains (mental health,
family relationships, substance use, etc.) summarized for each case. Two study team members (one of whom
was the primary interviewer) read the assembled case file data. Through consensus-building discussions, we
focused on each individual’s trajectory before drawing larger cross-case conclusions.

The term case study refers to the process as well as the outcome (Patton, 2002). The key distinguishing feature
of case study processes is that they maintain the holistic integrity of the case, that is, it is unpacked and its
contents closely examined, but the parts are ultimately viewed as a whole and in relation to one another. A
“case study-as-product” is a comprehensive description built up from immersion in multiple sources of data. It
brings in specifics but does not get bogged down in detail. Many of the same methods for thematic analysis
and pattern recognition apply to multiple case study analyses. As a rule, similar cases are easier to analyze
comparatively than dissimilar ones.

The end product of a case study analysis may be thematic, but it can also be typological. Take, for example, a

229
hypothetical cross-case analysis of eight child welfare agencies and their responses to new regulations
mandating professional training in social work for all staff. Each organization has its own professional culture,
leadership style, caseload mix, and staff morale. Based upon in-depth analysis of multiple data sources, the
organizations’ responses to the mandate tend to fall into one of three categories: earnest adopters (strong
leadership), active resistance (strong leadership), and passive resistance (weak leadership). Of course, one
should not force categorizing onto the data—cases that do not fit call into question the viability of the
typology. Yet the development of typologies—when relevant—is useful heuristically for multiple case study
analyses.

230
Data Analysis in Narrative Approaches
Individual narratives and conversations are revelatory events when analyzed for their cadences, interruptions,
intonations, emphases, and lyrical storytelling. Data analyses in narrative approaches draw on literary
traditions while reflecting a preoccupation with social and cultural meaning. Narrative data analyses fall
roughly into two types: (1) analyses of storytelling (narrative analysis) and (2) analyses of conversational
exchanges (discourse or conversation analysis). Under the expansive umbrella of narrative approaches,
differing procedures for data analysis can be found.

Following Labov and Waletzky (1967), traditional narrative analysis (NA) involves identifying six elements of
a fully formed narrative: abstract (summary or précis of the event), orientation (time, place, participants,
context), complicating action (what actually transpired), evaluation (meaning and significance of the event),
resolution (conclusion of the event), and coda (giving closure by returning the listener to the present time). A
couple of caveats are pertinent here. First, not all narratives contain all six elements, and analysts may
reasonably disagree about what constitutes a coda, evaluation, and so forth. Second, narrative stories may be
embedded within a long interview, or they can emerge over a series of interviews (Riessman & Quinney,
2005).

Clandinin and Connelly (2004) recommend three analytic axes for studying narratives: personal and social
interaction, continuity (past, present, and future), and setting or situation (of the story or narrative event). The
second of these points to the significance of chronological sequencing within and across narratives. Stories are
natural places to talk about life-as-lived. Narratives may be the building blocks upon which a larger thematic
analysis is conducted, or a single narrative chunk may be dense in meaning such that it is stand-alone.

Riessman (2008) distinguishes between narrative analyses of content (what is being said), structure (how it is
being said), dialogic/performance (focusing on the dialogue or to the participant performing the narrative, and,
images (using photographs or other visual media along with the spoken word). The interpretive turn in
qualitative methods shows in these expanded categories.

Thus, NA can take a number of forms ranging from microstructural analyses of texts to examinations of the
larger context of the storytelling experience (Riessman, 2008). With respect to the Labov and Waletsky “six
elements” approach, Riessman (2008) cautions “not all stories contain all elements, and they can occur in
varying sequences” (p.3).

One variant of NA examines the performative aspects of storytelling (Langellier & Peterson, 2004; Riessman,
2008). Echoing Goffman’s (1959) work on presentation of self, narrativizing is treated as a dramatic act
engaging the teller with the audience. Researchers may examine how respondents “voice” themselves and
others, thereby indicating social relationships and the meanings attached to them (Sands, 2004). The act of
telling, while having therapeutic effects in itself, is a window into personal identity construction and
reconstruction. Emplotment, or the stringing together of events into a narrative arc, creates a structure or
scaffolding. Illness narratives are among the most popular exemplars of NA (Charmaz, 2014; Mattingly,

231
1998).

Analyzing naturally occurring speech can uncover social and cultural influences that structure human
interaction. Conversation analysis (CA) does this by examining turn taking as well as silences and nonverbal
utterances that signal gender, age, and race (ten Have, 2014). Box 6.2 demonstrates the detailed system of
notation used in CA.

Discourse analysis (DA) includes texts as well as conversation. Analytic procedures center on spotlighting how
larger social influences (especially unequal power and dominance) shape modern discourse (Gee, 2005). In
conclusion, narrative approaches share a goal of understanding in the context of talk and verbal interaction.
Whether following the arc of a personal narrative or the back-and-forth of a conversation, the analyst’s job is
to extract meaning (Hyden & Overlien, 2004).

232
Box 6.2 Notation in Conversation Analysis: A Detailed Breakdown of
Text
Some of the marks used to transcribe conversations for CA are the usual punctuation marks such as the question mark (?) for rising
inflection, the period (.) for a full stop, and the comma (,) for a pause or in a listing. Below are specific additions used in CA
transcription:

(.) = a micropause
(2.3) (.6) = a 2.3 or 0.6 second pause (the time lapse of the pause)
[ ] = overlapping speech (brackets used to show where speakers are talking over each other)
_____ = emphasis (underlining used when the speaker’s tone is more emphatic)
CAPITAL LETTERS = loud (when the speaker’s voice becomes louder)
↑↓ = indicate lowering or rising pitch
: = drawing out the syllable before
° =degree sign used to surround quiet speech
hhh = out-breath
.hhh = in-breath
− = hyphen marks cut-off of preceding sound
> < = arrows surround speeded up talk
= equal sign refers to run-on sentences (within or among speakers)

Example 1: “From that day on he has nev::er put his hand up to me or at me. hhh (.)” This translates as the speaker drawing out the
word never, then comes to full stop, then exhales at the end of the statement and has a micro-pause in speaking.

Example 2: “He tries sometimes to be intimidating; he can’t do that (.5)” This translates as emphasis is put on that followed by a 0.5
second pause.

233
Phenomenological Analyses
Phenomenological analyses (PA) can take different forms, but they share certain features. These include
synopses of each SP’s experiences and a summary of major themes with associated excerpts from the
interviews (Giorgi, 1985; Moustakas, 1994). Moustakas recommends the researcher begin by recording his or
her personal experiences with the phenomenon to help set boundaries on limiting (but not eliminating) the
influence of those experiences. The concept of epoche is used to describe this distancing or bracketing of
personal experiences (Moustakas, 1994).

The next step, termed horizontalization, is to read across the interviews repeatedly to identify significant
statements in the data and group these into themes. Each statement is given equal value. Textural description
(Moustakas, 1994) is attached to these themes such that participants’ lived experiences are visible. In the next
stage of analysis, the researcher goes from what to how through structural description (a broader examination
of the context of the lived experience). For example, a phenomenological study of doctors addicted to
painkillers might describe how and where the addiction began. Were they going through a personal life crisis
or a serious medical condition? Did another physician provide the opportunity and encouragement? Did the
drug use begin outside of the clinic or during work time? The final step in PA is the synthesis of meanings
into essences. The essence is the condition or quality without which a thing would not be what it is. This is
done by using both types of description to arrive at a composite or blended portrayal of the phenomenon
(Moustakas, 1994).

Van Manen (1990) positions the researcher within the study as engaging in the data more holistically,
selecting and highlighting important statements for later aggregation. A step-by-step example of this is
provided by Groenewald (2004) in his phenomenological study of educational programs in South Africa.

234
Action and Community-Based Participatory Research
Action research and community-based participatory research (CBPR) emphasize cooperation in data
collection and analysis. The specific tasks follow the lead of the particular approach being used with the
proviso that community partners play a key role in analytic decisions. If time is limited, the more expedient
template approach (Crabtree & Miller, 1999) may be used in which key domains structure the data collection
and the analyses. This limits the inductive nature of the study but at the least answers the pertinent questions
in an efficient way. While some community partners may be content to leave the analyses to their research
partners, this does not exclude them from consultation at each step.

235
Analysis in Longitudinal Designs
Few qualitative studies are longitudinal in the sense of prospectively collecting data over long periods of time
(Flick, 2004), and qualitative methods have offered relatively little guidance in this regard (Saldana, 2003). A
number of qualitative approaches are defined by the amount of time needed to carry out the study and the
desire to analyze change in some way, so much depends upon the researcher’s intentions and resources. An
ethnographer may spend months or years in the field yet be less concerned about change than about thick
description. Similarly, a researcher could study a fast-moving set of events (the aftermath of a disaster) and
focus on change taking place in days and weeks. The study may have specified waves of data collection at
preset intervals or it may take place in situ over periods of months or years.

The analysis of longitudinal data follows the principles of qualitative analyses in general—it begins early and
follows an iterative course. What is distinct is the assumption that change and process are important to the
study (hence the longitudinal design). This usually means that the same or similar questions will be posed over
time to lay a foundation for temporal contrasts. A key decision point is whether one is coding and aggregating
versus keeping the case intact. If the former is chosen, one might code the baseline data then compare and
contrast baseline codes with the codes that emerge from subsequent waves. The case study option points to
multiple case analyses in which case summaries and other case-based data are analyzed over time to detect
trends and types of change. Deep immersion without demarcated data collection points requires keeping track
of change through extensive documentation and memos.

Life course studies and life histories typically depend on retrospective recall (with the occasional exceptions of
prospective longitudinal designs that are almost always quantitative) (Singer, Ryff, Carr, & Magee, 1998).
Qualitative studies bring out the dynamic, interactive aspects of lives-as-lived recalled by SPs. As such, they
depend heavily on the researcher’s ability to “draw the threads” yet maintain the integrity of the individual
trajectory—a task similar to cross-case analysis. Despite these challenges, such studies are vital to
understanding the full array of influences on individuals as they mature but also the interplay of their lives
with historical events and social change (Elder, 1994).

Of course, longitudinal does not have to refer to a lifetime. As mentioned earlier in this chapter, the NYRS
involved four waves of interviews over an 18-month period, and we elected to do multiple case study analyses
of individual trajectories based upon indicators of mental health recovery including mental health, use of
substances, social support, work or school involvement, housing stability, recreation/leisure activities, and
physical health. For each of the 38 individuals who completed all 18 months of the study, we had two team
members independently read the interview transcripts and case summary then meet to consensually fill out a
matrix display based upon the recovery domains. This matrix—a very large spreadsheet—allowed us to read
across to see how each individual was doing. We could also read down to look for commonalities across the
domains, for example, use of drugs or alcohol. The next step in the analyses was pattern recognition across the
individual trajectories to detect trends toward (or away from) recovery and what influenced those outcomes
(Padgett, Smith, Choy-Brown, Tiderington & Mercado, 2016).

236
237
Analyzing Visual Data
As qualitative researchers expand their data collection to incorporate visual media, new analytic opportunities
and challenges arise. Visual data analysis has become the subject of many works bridging the arts and social
sciences including a large edited volume (Margolis & Pauwels, 2011). The “language of pictures” (Noth, 2011,
p. 299) blends symbolism with images. Thus, a photograph has content (scenes, people, objects, etc.), but it
also has meanings filtered through personal predilections as well as cultural and social expectations. The
positioning of objects within the frame, the symmetry or asymmetry of the images, use of foregrounding and
backgrounding, and center versus periphery placements all convey information.

Visual data tap into emotional reserves in a unique way. The human brain more readily processes visual
images than spoken or written language. Neuropsychologists have found that the brain processes pictures and
words in differing ways, with the former having more immediate emotional impact (Kim, Yoon, & Park,
2004). According to LeClerc and Kensinger (2011), “For pictures, the effect of emotion might be in evidence
immediately and might be evoked relatively automatically, whereas activation of emotional responses for word
stimuli may require more in depth and controlled processing” (p. 520).

Researchers can develop typologies of visual images based upon participants’ narratives or a prior conceptual
schema. An inductive approach that we took in photo-elicitation interviews in the NYRS was to identify two
broad categories of the stance to photographing taken by participants. These were “a slice of life” and “then
versus now,” the former capturing everyday activities and the latter using the photos to narrate how different
their lives were compared to earlier days of homelessness and addiction (Padgett, Tran Smith, Derejko,
Henwood, & Tiderington, 2013). Later, we added a typology to distinguish “documentation” photos from
“representation and evocation” photos to take into account the accompanying narratives. A photo of a subway
stop could be narrated simply as the participant’s local station. Or it could form the backdrop to a discussion
of getting around the city and mobility (or of the noise, danger, and discomfort of subways) (Tran Smith,
Padgett, Choy-Brown & Henwood, 2015). In a final interpretive step, we noted how the act of taking and
narrating photos created an opportunity for “identity work” (Snow & Anderson, 1987; Tran Smith et al.,
2015).

Figure 6.2 displays a portion of a photo-elicitation interview transcript. The SP (#222) had taken the photo at
a community health clinic where he received medications for HIV infection. The quilt on the wall—a
memoriam to those who have died of AIDS—prompted him to share a run of emotions, including anger,
sadness, and gratitude at how welcome he felt at the clinic. We classified this photo as “representation and
evocation” because he used it to represent his health care needs being met, but he also embarked on a broader
discussion about his lack of health insurance, the grief of so many deaths, and his good fortune in being alive.

Figure 6.2 Screenshot From ATLAS.ti: Excerpt From Photo-Elicitation Interview in the NYRS

238
Visual data offer a degree of multidimensionality that lends itself to creative interpretations and analytic
decisions. When generated by the researcher, it adds depth to the corpus of data being collected. When
participants produce and interpret the visual data, it not only empowers them but it yields a rich dialogic
encounter with the researcher.

239
Coding
Most qualitative researchers who use coding stay with description and interpretation and stop short of theory
generation. Flick (2004) asserts that requiring theory development constitutes an excessive and unrealistic
burden for many studies, especially graduate theses and dissertations. This, of course, does not preclude
theoretical thinking or theorizing (which should be brought to bear early and often).

Coding qualitative data is the most commonly used analytic procedure as it produces the building blocks or
scaffolding for the study’s interpretive findings (Morse, 2015). That said, there is tremendous variety in how
coding is carried out (Saldana, 2015). For example, there are coding techniques independent of GT, including
content analysis and methods developed by Boyatzis (1998) and Patton (2002). The boundaries between GT
and these other methods are frequently blurry, but here are a few distinguishing features of GT. First, as
described in Chapter 2, GT is most often used to describe a process or action. Second, in GT methods, “the
ultimate goal of developing theory is never out of the researcher’s mind” (Oktay, 2012, p. 53). This means that
themes or categories are not the final product—exploring relationships between them is necessary to produce
even a “small t” grounded theory. Third, data collection and analyses are iterative and overlapping, and new
data collection using theoretical sampling is common.

240
Varied Approaches to Coding and Co-Coding
Coding breaks the “data apart in analytically relevant ways in order to lead toward further questions about the
data” (Coffey & Atkinson, 1996, p. 31). As noted by Tesch (1990), each chunk or quotation has two contexts,
one its origin in the narrative and the other a “pool of meaning” located in higher levels of abstraction. Coding
sets the stage for interpretation and is interpretation (albeit at a rudimentary stage). At the outset, one begins
with close and repeated readings of the transcript (or other text) in search of “meaning units” that are
descriptively labeled so that they may serve as building blocks for broader conceptualization. Somewhat like a
funnel, coding starts at a broad descriptive level and gradually contracts into greater selectivity and interpretive
synthesis. Meaning units often consist of events or incidents and the personal reflections that come with
talking about them. Participants’ recollections are coded for their meaning with an eye to broader
considerations.

Questions and decisions arise early on, for example, where code labels come from, to what extent do these
labels draw on a priori concepts and theories, and the level of detail attending the analysis (think of a fine-
versus a coarse-toothed comb). The answers to these questions depend on the study’s overall conceptual
framework and design. To be avoided across the board is a mechanistic approach that deprives the study of
new insights and interpretive thinking.

Co-coding, or the independent coding of the same data by two or more researchers, is valuable, as leaving this
important task to one person runs the risks of veering off course or getting stale after awhile. It is also central
to team research and community-based studies where members of the team share in research tasks. During
the early stages of open coding, co-coders are likely to come back with different codes and definitions even
after being oriented to the study and its goals. This is not a problem—diverse viewpoints are valuable. It does,
however, result in a rather messy and time-consuming process of talking through the coding decisions and
reaching consensus.

Some more positivist-minded qualitative researchers seek to quantify inter-coder agreement through calculating
Cohen’s kappa or another coefficient of agreement (Boyatzis, 1998; Mayring, 2004). If done, this should wait
until the codebook is finalized, all co-coders are trained in its use, and the rules for what constitutes
“agreement” fully understood. Even when coders agree on the label for a particular passage of text, they often
disagree on how much or little to assign to it. I am not a fan of calculating coder agreement as I believe it
requires many preparatory steps and borders on mechanistic. It also implies a degree of precision that can
squeeze out (or at least undermine) larger interpretations.

241
Starting Out: Identifying Code-Worthy Material
Most qualitative researchers begin with naïve or open coding (Charmaz, 2006; Ryan & Bernard, 2000). For
the novice, this can seem like working the trapeze without a net, but one need not approach coding as a blank
slate (pardon the mixed metaphors here). Grounded theorists refer to sensitizing concepts (Glaser, 1978) as
providing initial guidance on where to start looking. Thus, a study of persons with schizophrenia would likely
consider looking for “stigma,” and one examining eating disorders might look at “body image.” Regardless of
whether sensitizing concepts are invoked, the researcher approaches the text with as few preconceptions as
possible and holds the ones she has lightly. Open coding serves another purpose besides organizing the data; it
draws in the researcher to the data and prevents him from reverting to preformed ideas or conceptions.

A qualified exception to this comes from studies where codes are pre-existing and form the bins that structure
the data analysis. As mentioned earlier in this chapter, a template approach (Crabtree & Miller, 1999) happens
when the data are forced into preselected codes. These codes might be derived from theoretical and
conceptual frameworks or from previous research. Though moving the study away from a major strength of
qualitative methods (inductive thinking), there are occasions when time is limited, and such structure is
necessary.

A template approach was imperative in an evaluation of foster care in New York City given the topic as well
as the limited time available (Freundlich, Avery, & Padgett, 2007). Consultation with key stakeholder groups
(youths, social workers, attorneys, judges) and the literature produced seven domains that structured the open-
ended questions as well as the analyses. These domains were youth involvement, transitioning,
recommendations for improved services, quality of placements, safety in the placements, services in the
placements, and permanency planning. Such prepackaged codes permit expedited data collection and analysis
yet their derivation stayed true to stakeholder opinions.

When open coding, one can use the right-hand margins of the transcript to bracket relevant segments and
assign code labels to them. Although this may be carried out directly on the computer screen using QDA
software, I prefer marking hard copies first (always using pencils with erasers). Flipping back and forth the
pages of a transcript is easier than maneuvering a cursor in a limited-view window (and is also easier to do in a
comfortable chair). Before and during coding, I use yellow highlighting when I come across interesting
quotes. It is not always easy to flag these, and some are not used in the write-up, but there is no greater
pleasure for me than coming across an eloquent statement that perfectly captures a code’s meaning.

A few important considerations arise at this point. First, every line of the transcript is not necessarily coded
(or code-worthy). Second, a single passage of text may be so rich that it yields several code-worthy chunks of
information. Coding can get messy in such instances—sometimes one must literally circle the relevant text
with an arrow leading to the code label in the margin. (Right-hand margins can look like traffic gridlock when
an interview yields a lot of important material.) Third, codes need to have clear definitions to guide their
usage, that is, what belongs in and what does not. Finally, codes are provisional and subject to change, either
through clarification and revision or outright elimination.

242
Early in data analysis—usually after three or four transcripts—a start-list of codes is compiled and applied to
additional transcripts. A commonly used approach is to have two persons independently code the first few
transcripts, then meet to discuss their findings and consensually agree on a provisional list of codes. In this
way, new codes may be added and excess codes discarded.

Trimming back a proliferating code list is inevitable. Codes get dropped for two primary reasons: (1) they
have too few excerpts (or their content is too thin), and (2) they become merged with or absorbed by another
code. A code’s staying power is not a matter of quantity but quality. After coding a few more transcripts (this
number varies depending on the density and richness of the data), the list of codes starts to saturate and no
new codes emerge. The size of the final list can vary considerably, but it tends to become unwieldy when
codes number more than 30 or 40. Less (or fewer) codes can be more.

Table 6.1 shows a portion of a transcript from the NYRS along with examples of open coding and brief
memos in adjacent columns (line numbers next to the codes show the chunk being labeled). Some of the code
labels are straightforward descriptors, for example, “loss of job,” and others pull in a phenomenon, for
example, “supporting the (drug) habit.” The use of gerunds suggested by Glaser (1978) can be seen in these
two examples, the first a discrete event that did not merit a gerund but the second an ongoing process
(burglaries and theft to support drug use) that did. The reader may note that some passages are dense enough
to be linked to multiple codes. In the memo column are brief “notes to self” while reflecting inward (to this
specific interview’s content and meaning) and outward (larger issues that might recur in other interviews).
Memos are an analytic conversation and in long form should be recorded in a separate journal or file. Box 6.3
displays what a longer memo looks like when applied to this SP. Call-outs or action notes in Box 6.3 are
underlined to alert us to additional work to be done stimulated by this interview.

243
244
245
 Box 6.3 Example of a Memo for Interview Excerpt in Table 6.1
SP is a middle-aged African American man who has spent most of his adult life in prison for drug-related crimes. He suffered from
a serious mental illness that went untreated for decades. Spending time in the “box” (solitary confinement) was, to say the least, not
helping him cope with his mental illness. The availability of drugs such as angel dust inside the prison seriously exacerbated the
violent behavior he attributes to the voices in his head. Action note: check on how common this is (illicit drug use in prison and long
stints in the box and whether it is related to untreated mental illness). Although SP does not use the term as such, his life reached a
turning point in 1999. In a later interview, SP reveals that he was diagnosed with schizophrenia while still in prison in 1999 and put
on medication. Does this mean his medication worked? Is there a link to his stopping drug use at about the same time? Action note:
Ask SP about this in next interview. Upon his release from prison, the SP went to a shelter, a place hardly conducive to positive
change, but he had already persevered in overcoming drug dependence and mental illness while in prison. Questions going forward:
Can we assume that the “box” was associated with a worsening of untreated mental symptoms? When illegal drugs are available in
prison, what strategies are needed to abstain? Is drug use a form of self-medication? After release from prison, how did SP manage
so well?

A few caveats immediately come to mind when regarding Table 6.1. First, these code labels and memos are
not the only way to go; another researcher might reasonably use others. Second, as shown by the line
numbers, decisions have to be made about how much material to grab. There are no hard and fast rules here.
Some researchers prefer more and others less. Finally, each code label must earn its way by having
representation in this and other interviews. A researcher knowledgeable about the population of study (and
prior research) is more likely to identify salient codes, but he is not infallible. I have returned to a coded
transcript and noticed some passages I overlooked or codes I could have invoked.

In the same manner, some analytic memos are more consequential than others—it all depends on the study’s
ultimate goals. For example, a researcher interested in mental illness and incarceration might put more time
and analysis into experiences of the “box,” or solitary confinement. There is a growing literature on this subject
and on the effects of mass incarceration in the United States (Alexander & West, 2012). For our study,
mental health recovery being the focus shaped how we appraised interview content. Thus, we were struck by
how the SP matter-of-factly attributed his improved mental health to his medications (other SPs resisted such
conclusions or cited side effects as superseding any positive effects). All of this is grist for the interpretive mill.

246
Labeling Coded Material and Developing a Codebook
Each researcher brings to coding an array of personal and professional experiences and a lexicon for naming or
labeling them. As a rule, one should resist the temptation to use professional jargon in labeling codes. Staying
close to the data helps prevent this from happening because few individuals express themselves this way in
normal conversation (e.g., referring to their “denial” or “coping style”). Code labels should be brief yet
descriptive. In the NYSS, we used the code “living independently” to refer to occasions when study
participants talked about the benefits of having an apartment of their own. The use of a gerund conveyed the
activities of apartment living—cooking, cleaning, relaxing—rather than a static expression such as “my own
apartment.”

Code labels should be compelling and interesting when possible. Statements like “I take my nephews to
school every day” and “I get medicines for my elderly neighbor” might be coded rather dryly as “altruistic
actions” or more evocatively as “helping others.” Less inductive are code labels that correspond directly to the
question that was asked. Thus, a code “next step” might be used to label participants’ responses to the question
“what is your next step?”

Code labels may be in vivo, emerging directly from participants’ words. In vivo codes are pithy comments
made by participants that have meaning beyond their immediate connotation. Sometimes an in vivo code is
the product of jargon or street talk, for example, study participants who talk about their struggles with
addiction might speak of “hitting bottom.” Used as a code label, this phrase describes a critical turning point
in one’s recovery. An in vivo code we used early on in our research (“I’m not like them”) was a phrase used to
convey a sense of being superior to or apart from others within the same treatment program, shelter, or
psychiatric hospital. This sentiment connoted frustration with being housed and treated alongside other
psychiatric patients who were perceived as more disturbed.

Code labels may come from the researcher’s knowledge of the literature, what Oktay (2012) refers to as
“theoretical” codes. Incurring a higher risk of forcing the data, theoretical codes should be used judiciously and
with good reason. Tiderington (2015) used the code “boundaries” to describe how case managers navigate
their meetings with clients and maintain professionalism. For social workers, “boundaries’ refer to clinical
theory regarding the therapist’s avoidance of personal disclosures and honoring the client’s privacy. Here is a
quote that fell under the code given by a case manager describing how she overlooks her client’s crack cocaine
use as long as he hides it during her apartment visits. “But please, when I come . . . you know I’m coming
because you know I have a schedule, please don’t have that on the table. Please put that stuff up! And they
respect that” (Tiderington, 2015, p. 86).

Table 6.2 shows a codebook developed during the NYSS. There is no standard format for this; every
codebook has its own logic and procedures. Below are a few aspects of this codebook worth noting.

247
248
Several codes capture responses to standardized open-ended interview questions. These were “current
life situation,” “homeless experiences,” “program experiences,” “substance abuse/use,” “social
relationships,” “thoughts about the future,” and “what’s missing.” The other codes were inductively
derived.
Some codes are divided into subcodes. For example, “housing experiences” was a topic about which we
probed extensively to understand participants’ experiences (e.g., the benefits of having housing) as well
as feelings (e.g., fears of losing their housing). In our experience, subcodes are inductively derived, a
reflection of the expanded meaning accorded a topic by participants. The broader a code’s presence in
the data, the more likely it is to be subcoded.
We abbreviated code labels, for example, “mental illness experience” became “MIE.” This made hard
copy coding easier and faster (a time-saving step not needed when pulling from a software code menu).
Each code or subcode had a definition that was both inclusive and exclusive to reduce coding errors, for
example, “self reflection” was defined to exclude “thoughts about the future” as this was a code unto
itself.
When a code represented an experience that could be positive or negative we used the shortened “+” or
“−” to show valence differences.
Some of the codes had relatively thin representation but were too important to leave out given the focus
of our research, for example, suicide, stigma, and work experience. Put another way, only a few
participants revealed suicidality, had work experience to talk about, or reflected on being stigmatized
(for being mentally ill, homeless, or addicted), but all of these rare occurrences were too important to
overlook.
Two of the codes proved to have too little grab and were ultimately discarded but remained alive
through memos. “Aging effects,” for example, was incorporated into our findings to connote maturation
and recovery from drug or alcohol abuse (Henwood, Padgett, Tran Smith & Tiderington, 2012).
“Turning point” was a code we identified early on, but it subsequently did not hold up, that is,
participants did not narrate their lives in this way. Yet we kept this idea in mind as we examined
meaningful life events in larger contexts and through a retrospective lens.

249
Documenting Coding Procedures
Coding is a profoundly discretionary activity; it involves painstaking sifting and sorting, but it is also
intellectually demanding. At the outset, it is not unusual to have dozens or even hundreds of codes. (When
two or more persons are involved in coding, one person needs to be in charge of the code master list to
maximize coordination.) To avoid descending into a black box obscuring all that happened, the researcher
should document coding procedures in a way that does not interfere with the inductive and serendipitous
course that coding takes.

In what Miles and Huberman (1994) refer to as choreography, this coordinated multi-tasking invites the use
of a data management log. In her dissertation research, Tiderington (2015) kept such a log with dates and
tasks accomplished. Iterative versions of coding decisions, sensitizing concepts used and dropped, consensus
meetings with co-coders, “breakthroughs” or insights that later yield themes, and meetings with her
dissertation advisor are all documented in the log (Tiderington, 2015). If a team of researchers is involved, one
person may keep the log or members may take turns.

Co-coding raises questions about how to deal with disagreement. As discussed earlier, statistical analyses are
available for calculating inter-coder agreement, but they impose a degree of mechanization on the coded data.
I prefer consensual validation in which co-coders meet to compare their results, relying on cogent arguments
and persuasion grounded in the data. That this process is organic and iterative ensures flexibility and openness
to change. Of course, it is untidy in the early stages, and disagreements are almost inevitable. Thinking about
codes, discussing and defending them, and reaching consensus are valuable aspects of the process, not
something to avoid or prematurely conclude. These activities honor the richness of the data and take into
account the inevitability of multiple standpoints. Documenting this process helps make it more transparent
and meaningful to those involved.

250
Comparing and Contrasting: Memo Writing as Forward Motion
Memo writing is central to the process in which one documents thoughts and ideas that emerge through
interacting with the data (Saldana, 2015). Strauss and Corbin (1994) distinguish between three types of
memos: code notes, theory notes, and operational notes. Code memos are the basis for definitional statements and
documentation of their reason for being. Theory notes are a record of ideas and hunches about what is going
on in the data. Operational notes are placeholders for logistical and other concerns. Whether one wishes to
categorize memos or not, they are safety zones for discovery and creativity, a place for hunches and conjecture.
They may be shared with one’s advisor or collaborators but otherwise should remain private repositories of
ideas.

As coding and data analysis proceed, the analyst remains cognizant of similar incidents in other interviews as
well as the larger context, searching for patterns but also remaining alert to negative instances and
irregularities. Keeping track of this with memos helps to propel forward the analytic process. Here is where
reflexivity surfaces again as critical. As the researcher dives deeper into the data, analytic isolation brings the
risk of going off-track, either through inviting in preconceptions or through getting lost in the weeds or
minutiae. Being mindful of this, it helps to periodically think through one’s role as immersed but still analytic.

What grounded theorists refer to as constant comparative analysis describes a systematic search for similarities
and differences across interviews, incidents, and contexts (Strauss & Corbin, 1994). A constant comparative
analysis stays close to the data, but its ultimate value comes from an ability to make sense of myriad
comparisons, winnowing through them to note what is meaningful. In practice, comparative analysis is
cyclical, beginning as inductive, then becoming deductive, and then returning to the inductive. This cycle can
be repeated many times over.

Memos evolve and become more focused as patterns are recognized and interpretations come into play. Being
part of a team, debriefing with another qualitative researcher or having a like-minded mentor can be extremely
helpful as one shares memos and thoughts, testing them out and refining them to stay close to the data but
also venture into higher-level abstraction and interpretation. As important as it is, coding is only the starting
point in data analysis. Subsequent cycles of coding help to narrow the analyses to ever-more important
findings.

251
Secondary Analyses of Qualitative Data
Given the sizable investments of time and resources that go into a qualitative study, it makes sense that many
qualitative researchers are willing to share their data with others. As a general rule, richer and complex data
offer more fertile ground for secondary analyses than narrow or thin data. Secondary analyses extend the life of
a study and are an efficient use of resources. Although they do not provide the warmth and good cheer that
follow a successful interview or day in the field, their availability is a viable trade-off (Williams & Collins,
2002).

Thorne (1998) distinguishes five types of secondary analysis beginning with analytic expansion, in which the
researcher ventures into new topics using her own data. Retrospective interpretation involves going back to the
data to further develop themes. Armchair induction applies new textual analytic techniques to existing data.
Amplified sampling involves comparing several distinct databases for broader analysis. Lastly, cross-validation
takes the researcher beyond her immediate findings to seek confirming or disconfirming evidence from other
databases. From a logistical standpoint, the analysis may approach the raw data for open coding or it may
work with codes developed in the earlier study.

The original research team may carry out secondary analyses, or some of the team’s members may collaborate
with new researchers. This is preferable to sharing data with a new investigator who has minimal knowledge
of the original study. The new researcher, for example, will not likely be able to return to study participants for
additional data collection nor would he have the same intimate knowledge of the data. Regardless of who is
involved, there is the potential problem of retrofitting new research questions to the earlier data (Thorne,
1998; Williams & Collins, 2002). Although usually more open and raw than quantitative datasets, qualitative
data have undergone their own filtering and sorting processes unknown to new investigators. The norms for
sharing memos and analytic decisions are far from agreed on within qualitative research, and many
investigators do not keep such documentation in the first place. Finally, the interconnectedness of researcher
and participant, one of the most rewarding and informing aspects of qualitative research, is missing from
secondary analyses. Qualitative researchers who prefer collaborative models of inquiry with their study
participants would find this unsatisfying.

On the positive side, gaining institutional review board (IRB) approval for secondary analyses is usually much
easier because interacting with “human subjects” is not involved. Yet, secondary analyses are not without
ethical risks. Identifying information, for example, can inhere in the data even when the names of participants
are not included (Thorne, 1998). The respect for privacy and awareness of context that characterized earlier
relationships may be missing on the second go-around. Data from a study of abortion among college students,
for example, could become the basis for interpretations that range well beyond what the study participants
believe they had signed on for in the first place. Such ethical concerns must take priority in determining with
whom and how qualitative data are shared. In addition, participants consented to the first study, but should
they be asked to consent to this second round? Consultation with one’s local IRB is highly recommended, as
there are no set rules for navigating this process.

252
253
Hybrid and Mixed Qualitative Approaches
As discussed in Chapter 2, researchers often choose to mix and match qualitative approaches. This can occur
at different stages of a study and unfold in differing combinations and for different purposes. It can also
introduce problems in the form of method and data incongruities (Johnstone, 2004; Wimpenny & Gass,
2000). As noted by Annells (2006), mixing at the analytic level can be problematic if the underlying
philosophical paradigms are in conflict—a lesser concern for pragmatists. Arguing in favor of hybrid vigor,
Fereday and Muir-Cochrane (2006) used inductive and deductive thematic analyses by mixing Boyatzis’s
(1998) inductive methods with Crabtree and Miller’s (1999) template style of coding to study nurses in their
day-to-day performance. Thus, data-driven codes from participant interviews, such as “trust and respect” were
combined with theory-driven codes such as “reciprocity” to produce an integrated model of performance
feedback and self-assessment in nursing practice.

An example of sequential mixing during analyses can be found in Beck’s (1993) study of postpartum
depression in which she conducted phenomenological analyses followed by grounded theory analyses of the
data. Close concordance was found between the two sets of findings, thus leading to the development of a
substantive theory explaining the onset and course of depression shortly after the birth of a child (Beck, 1993).
Agar and MacDonald (1995) used juxtaposition to compare the results of conversation analyses of focus group
data where teens discussed drug abuse with their ethnographic observations of adolescent drug use. An
important caveat applies here: Mixing can lead to conscious or unconscious blurring of the analytic
approaches. Keeping intact the analytic procedures takes a degree of vigilance to ensure each is making a
contribution.

254
Summary and Concluding Thoughts
This chapter began by offering guidelines for managing and analyzing data and strengthening the
development of conceptual schemas via negative case analysis, theoretical sampling, and other strategies.
Computer software facilitates these activities in a number of helpful ways, but it cannot analyze the data.

Despite an array of possibilities, most qualitative data analyses have the following in common: (1) full and
repeated immersions in the data; (2) going “deep” into descriptive specificity as well as “across” with pattern
recognition; (3) attending to context—temporal and environmental; and (4) proceeding “up and out” to weave
in theoretical and empirical knowledge from the literature (more on this in the next chapter). Analysis begins
inductively, but the pathways to its completion often include deductive thinking as well. The insider
perspective is an invaluable part of this process, but the ultimate contribution of a qualitative study depends on
the probity and intellectual clarity of its interpretations.

Each of the six qualitative approaches has its own analytic traditions, yet the emphasis in this chapter was
given to what is done most often—coding and thematic development. Whether this leads to a fully developed
grounded theory or, more commonly, an interpretive thematic framework, such analytic involvement is an
exercise in restraint (from becoming weighed down by a priori ideas and concepts) and creativity (comparing
and contrasting, searching for what is unusual and unknown). Despite its demands, the initial stages of
qualitative data analysis are an exciting and necessary step to making previous efforts come to fruition.

255
Exercises
1. Conduct and audiotape a brief (15-minute) interview with a colleague or classmate asking for his or her
most memorable childhood experiences (good or bad). Transcribe the audiotape. How long did it take
to transcribe? Were there vocal utterances or other sounds during the interview that were meaningful
but not captured in the words?
2. Take the first five pages of your transcript from Exercise #1 and open code it following the instructions
in this chapter. At the same time, write two to three memos on what you are hearing and what might be
going on with regard to childhood experiences and their contexts.
3. Go to a qualitative journal listed at the end of Chapter 1, and locate a study that uses a noncoding form
of analysis. How are the methods described? How are the findings presented?

256
Additional Readings
Bernard, H. R., & Ryan, G. W. (2009). Analyzing qualitative data: Systematic approaches. Thousand Oaks,
CA: Sage.

Charmaz, C. (2014). Constructing grounded theory: A practical guide through qualitative analysis (2nd ed.).
Thousand Oaks, CA: Sage.

Coffey, A., & Atkinson, P. (1996). Making sense of qualitative data. Thousand Oaks, CA: Sage.

Dey, I. (1993). Qualitative data analysis: A user-friendly guide for social scientists. New York, NY: Routledge.

Dey, I. (1999). Grounding grounded theory: Guidelines for qualitative inquiry. San Diego, CA: Academic Press.

Flick, U. (Ed.). (2013). The Sage handbook of qualitative data analysis. Thousand Oaks, CA: Sage.

Gee, J. P. (2005). An introduction to discourse analysis: Theory and method. London, UK: Routledge.

Grbich, C. (2012). Qualitative data analysis: An introduction. Thousand Oaks, CA: Sage.

Harding, J. (2013). Qualitative data analysis from start to finish. Thousand Oaks, CA: Sage.

Hsieh, H., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative Health
Research, 15, 1277–1288.

Lofland, J., & Snow, D. A. (2005). Analyzing social settings: A guide to qualitative observation and analysis.
Belmont, CA: Wadsworth.

Miles, M. B., & Saldana, J. (1994). Qualitative data analysis (3rd ed.). Thousand Oaks, CA: Sage.

Moustakas, C. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage.

Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

Riessman, C. K. (2008). Narrative methods for the human sciences. Thousand Oaks, CA: Sage.

Saldana, J. (2012). The coding manual for qualitative researchers. Thousand Oaks, CA: Sage.

Silverman, D. (2015). Interpreting qualitative data (5th. ed.). Thousand Oaks, CA: Sage.

Stake, R.E. (2005). Multiple case study analysis. Thousand Oaks, CA: Sage.

Strauss, A. L., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques.
Newbury Park, CA: Sage.

ten Have, P. (2007). Doing conversation analysis (2nd ed.). Thousand Oaks, CA: Sage.

Thorne, S. (2000). Data analysis in qualitative research. Evidence-Based Nursing, 3, 68–70.

257
Weston, C., Gambell, T., Beauchamp, J., McAlpine, N., Wiseman, C., & Beauchamp, C. (2001). Analyzing
interview data: The development and evolution of a coding system. Qualitative Sociology, 24(3), 381–400.

Yin, R. K. (2013). Case study research: Design and methods (5th ed.). Thousand Oaks, CA: Sage.

258
QDA Software Resources

259
Readings
Bazeley, P., & Jackson, K. (2013). Qualitative data analysis with NVIVO. Thousand Oaks CA: Sage.

Drisko, J. W. (2004). Qualitative data analysis software. In D. K. Padgett (Ed.), The qualitative research
experience (pp. 189–205). Belmont, CA: Thomson.

Friese, S. (2014). Qualitative data analysis with ATLAS.ti. Thousand Oaks, CA: Sage.

Gibbs, G. R. (2007). Media review: ATLAS.ti software to assist in the qualitative analysis of data. Journal of
Mixed Methods Research, 1(1), 103–104.

Kelle, U. (1996). Computer-aided qualitative data analysis: Theory, methods, and practice. Thousand Oaks, CA:
Sage.

Lewins, A., & Silver, C. (2007). Using software in qualitative research: A step-by-step guide. Thousand Oaks,
CA: Sage.

Weitzman, E., & Miles, M. (1995). Computer programs for qualitative data analysis: A software sourcebook.
Thousand Oaks, CA: Sage.

260
Software Information
An excellent independent resource (not supported by a software company) can found on the Web at
http://caqdas.soc.surrey.ac.uk

261
Products
General product information: http://www.scolari.com
ATLAS.ti: http://www.atlasti.com
Dedoose http://www.dedoose.com
The Ethnograph: http://www.qualisresearch.com
HyperRESEARCH: http://www.researchware.com
NVivo: http://www.qsrinternational.com
QDA Miner: http://provalisresearch.com/products/qualitative-data-analysis-software
QDAP: http://www.umass.edu/qdap
QUIRKOS: http://www.quirkos.com

262
7 Interpretation Recognizing Salience

Devoting a separate chapter to interpretation is testimony to the artful complexity of this almost-final stage of
the qualitative journey. It is also an indication of how important—yet often neglected—interpretation is in
qualitative research. Benign neglect is understandable. There are no strict rules, nor should there be. But there
are methodological traditions to respect and paradigms to consider. And there are many productive pathways
to interpreting qualitative data.

Interpretation is integrative and syncretic. It is also risk: the risk of self-interest peaking through, the risk of
underwhelming insights, or the risk of overreach. Multidisciplinary perspectives are critical for offering fresh
ideas and framing devices. As with a kaleidoscope, different and equally fascinating combinations of ideas and
insights can come together depending upon the researcher’s fund of knowledge and creative thought
processes. Indicating that interpretation goes beyond idiographic description does not denigrate the latter, but
it does confer a necessary-but-not-sufficient status to description.

This chapter revolves around time-tested ways to enhance the quality of interpretation. The tie that binds
them together is salience. Defined as the perceptual quality that lends meaning to something, salience is
manifested when the data analyses yield rich and relatable findings. Interpretation is the act of explaining
what is salient and why.

As with all other phases of a qualitative study’s design, interpretation is iterative and achieves greater focus
and clarity over time. In what Tesch (1990) refers to as decontextualizing and recontextualizing, analyses of
data implicate pattern recognition, drawing on comparisons and contrasts. Keeping the momentum going, the
researcher circles back to the study’s original research questions and theoretical lenses as well as the literature
—keeping in mind what is known and not known.

Interpretation entails building up and out. Building up happens when moving from coding or other
foundational analyses to ever-higher levels of abstraction. Building out brings in the literature and incites
theorizing. Memo writing is important in fleshing out and documenting ideas. Over time, the memos help
move the study to its culmination. Although the final product may range from a constructivist-inspired
performance to a fully realized grounded theory, the vast majority of qualitative studies converge on some
form of narrated synthesis of themes and conceptual framings.

263
The Role of the Researcher(s)
The role of the researcher in interpretation pivots on stance. As we have seen throughout this book, stance is
the product of one’s epistemological and paradigmatic leanings. It can range from detached observer to fully
engaged co-participant, from believing that some form of reality can be captured to a belief that all reality is
socially constructed. Clifford Geertz (1973) assumed the position of “knowing outsider” in writing about the
deeper meaning of Balinese cockfights, but these were his interpretations, not those of the Balinese people he
lived among. As noted by Charmaz (2014), Erving Goffman’s brilliant insights into stigma and spoiled
identities were refracted through the sensibilities of a White privileged male living in the mid-20th century.
We are invariably influenced by the historic era, disciplinary norms, and interpretive trends that surround us.

These days, omniscience as a stance is not a desirable option. For constructivist researchers, the optimal
solution is to co-create with one’s participants. For action researchers, such co-creation serves empowering
goals shared by researchers and community members working collectively. Clearly, if there is more than one
researcher on the team (or if a solo researcher is working with participants co-equally), the choice of stance is a
negotiated one. What unites qualitative researchers embarking upon interpretation is the return to reflexivity
and mindfulness.

A risk at this (and any stage) of qualitative interpretation is that of bringing in one’s unexamined
preconceptions and assumptions. The emphasis on unexamined here is a necessary distinction acknowledging
that such preconceptions exist and are not problematic as such—the key is what one does with and about
them. In anthropology, the exposure of such preconceptions became a cottage industry after the 1970s as the
works of leading ethnographers were questioned for their presumed objectivity and the absence of reflexivity
(Van Maanen, 1988). For some, this inward gazing became an end in itself and auto-ethnographies and
personal memoirs supplanted learning about the unfamiliar.

The infiltration of values in research is a given these days. A commitment to social justice will likely sensitize
the researcher to inequalities and oppression during interpretation. This is not a problem unless it produces
inauthentic results or unreflexive analyses.

264
Framing Devices to Assist in Interpretation
What are some ways that qualitative researchers approach their data to ensure conceptually rich
interpretation? Below are nine strategies developed by Bohm (2004, p. 273) and inspired by the work of
Barney Glaser. I have taken the liberty of dividing them into two basic types as follows:

I. Change-oriented studies examine

1. Processes (phases, transitions, sequences)
2. Strategies (tactics, techniques, mechanisms)
3. Turning points (critical junctures, point of no return)
II. Nonchange-oriented studies examine
1. Interactions (mutual effects, interdependence)
2. Degree or intensity
3. Typologies
4. Identity (self-concept, self-reflection)
5. Cultural and social norms
6. Consensus (conformity versus conflict)

The above is a menu of options and the bipartite categories an imperfect means of injecting temporality
(imperfect because analyses under Part II may also be change-oriented). It is probably obvious that several are
aligned with a particular qualitative approach or disciplinary perspective. For example, #1 is the sine quo non of
grounded theory, and #8 underlies ethnography. Social and political scientists are likely to draw on #2, #3, and
#9, and psychologists will relate to #2, #4, and #7. Social work researchers can use any of these to productive
ends, but they also have interests that are likely to influence the interpretive phase. These include the
following:

The tension between individual agency and larger structural forces

Social justice values and power differentials
Intersectionality and multiple identities or standpoints
Systems thinking (families, communities, social structures)
Ecological perspectives (interconnections of person and environment)

For any given study, several of these framing devices are likely to stimulate thinking and help shape
interpretation. And, whether acknowledged or left tacit, they underlie much of the qualitative research found
in the literature today.

265
Using Qualitative Data Analysis (QDA) Software for Interpretation
As mentioned in Chapter 6, QDA software provides an efficient means of working with data. QDA software
packages also have functions for analyzing nontext data (images, audio, and video) and for inserting hyperlinks
to connect instantly to other files or data sources. In addition to filtering by codes or other labels, one can
select subsets of data according to participant characteristics such as sex, age, and the like. For example, men
and women might be compared in their responses coded under “sexual experiences.” The points of
comparison function as analytic axes.

A useful interpretive function is the ability to create concept maps and plot horizontal and/or vertical
connections between the nodes (e.g., “diet” and “exercise” can be connected to the higher order code “self-
care”). These flexible network-making utilities make it possible to display static relationships as well as flow
charts. Among the various types of software, NVIVO has the most developed capacity for theory development
and supports the creation of hierarchical “tree structures.” It is also the most demanding to learn. In general,
QDA software is better suited to grounded theory analyses (Kelle & Erzberger, 2004).

The popularity of QDA software increases the potential for mechanizing analyses and casting a gloss of
technological precision over an untidy, iterative, and painstaking process. Similar to quantitative statistical
software, it is all too easy to fall into “fishing” or “data dredging.” Most importantly, QDA software cannot
compete with the human brain. It makes organizing and visually scrutinizing data much easier, but it cannot
replace the human capacity to juggle many thoughts and ideas simultaneously.

266
Saturation
The term saturation refers to the point at which no additional data collection is needed, and the analytic
framework is fully fleshed out (Bowen, 2008; Morse, 1995). Studies with modest aims and a priori codes are
likely to be saturated much sooner than more ambitious and inductive endeavors. Saturation does not depend
on counts and frequencies because fullness comes from depth rather than breadth. Perhaps not surprisingly, it
is easier to assert having reached this stage than to prove it, although careful documentation can alleviate some
concerns about premature closure.

Box 7.1 offers an example of a study documenting how saturation was achieved through quantification. The
authors (Guest, Bunce, & Johnson, 2006) conclude that fewer interviews were needed than anticipated to
reach the saturation point, but they also caution that predicting this number in advance is highly risky.

267
 Box 7.1 Operationalizing Saturation: A Study of Sex Workers in West
Africa
Guest, Bunce, and Johnson (2006) decided to explore the question of saturation empirically using data analyses from their qualitative
study of sex workers in West Africa. When, they asked, would the codebook be complete and no further changes needed? Would six
interviews yield as much information as 12, 30, or 60 interviews? Could coding activity slow down to a trickle after X number of
interviews then suddenly yield new codes much later after “saturation” had presumably been achieved? Starting with 60 interviews,
they began coding in increments of six, keeping track of the number of codes and code definitions as they proceeded. Here is what
they found:

After coding the first 30 interviews, the codebook contained 109 codes of which 80 (73%) were identified within the first six
transcripts.
Of the remaining 29 codes, 20 were identified in transcripts 6 through 12 (now up to 90% of all codes).
Coding of transcripts 31 through 60 yielded only five new codes; four of these were spin-offs from existing codes rather than
entirely new information.
The strength or reach of a code was measured by the number of interviews in which it occurred rather than the number of
coded excerpts (because talkative interviewees could easily skew the data). Of the 36 high-frequency codes, 34 had been
identified within the first six transcripts.

Guest and colleagues concluded that 12 interviews would have been sufficient (rather than the 60 they conducted). However, they
offered a few caveats about this finding. First, they used a standardized list of open-ended questions about beliefs and experiences
surrounding sex work. Studies with more improvisational questioning would not be likely to yield such a stable list of codes early on.
Second, the coders had similar backgrounds and knowledge about the phenomenon so that consensus—even when coding
independently—was probably easier. Put another way, there were fewer “lumper versus splitter” disagreements that often beset co-
coding comparisons. Finally, their sample was relatively homogeneous. Response variation led more often to definitional tweaking
than to coding overhauls (dropping codes, adding new ones, etc.).

Commentary: Qualitative researchers naturally fear that premature closure of data collection will deprive them of vital information
waiting just around the corner. Yet unnecessary interviews and diminishing returns drain scarce resources and burden respondents.
The study done by Guest and colleagues (2006) is innovative for taking on the “saturation threshold” as an empirical question. Yet
its findings cannot be generalized too far because they were based on several restrictive premises that do not fit many qualitative
studies.

In a larger sense, saturation is a useful heuristic applicable to all of the qualitative approaches. Rich and dense
data will take longer and yield more; thin data will yield less. At the same time, invoking saturation should be
done carefully, as its accomplishment is largely a matter of trust in the researcher’s decisions.

268
Bringing Theor(ies) in
As noted in Chapter 1, theories play a role in qualitative research but not a presumptive one—the grounded
theory concept of theoretical sensitivity comes closest to capturing this attunement to existing theories while
ensuring that they do not dictate the terms of the analysis and interpretation. Chapter 1 also introduced
alternatives to grounded theory that treat theories not as end product but as an analytic partner. Analytic
induction, for example, brings in theories (or explanations) to explain cases in an iterative fashion, looking for
the best match of the evidence with the theory (Bernard & Ryan, 2010). Abduction starts with anomalous
cases or perplexing problems and moves through a process of hypothesis framing and reframing meant to
produce the most parsimonious explanation (Timmermans & Tavory, 2012).

For good or ill, inductive thinking only rarely results in theory generation or theory fitting. Qualitative inquiry
has its origins in rejecting the outsize ambitions of 20th-century theories. This does not mean that theoretical
thinking is rejected (far from it), but such thinking is rooted in local knowledge and has more modest aims.
Take-away messages are often left up to the reader rather than predigested. Alternative ways of thinking
about the desirable outcomes of a qualitative study include verstehen (deep understanding) and thick
(interpretive) description.

More compatible with qualitative inquiry are open systems theories such as symbolic interactionism (Blumer,
1969), ecological theory (Bronfenbrenner, 2004), and Freire’s emancipatory approach (1973). Concepts—
drawn from theories or having enough heft to stand alone—are also drawn back in. For example, Bourdieu’s
(1999) concept of “habitus” obliges the researcher to “have knowledge of objective conditions” affecting
participants’ lives beyond the interview setting (p. 613). Taking things said at face value and going no further
undermines the researcher’s ability to make larger points, especially about institutions and structures that
hinder social progress.

In the New York Services Study (NYSS), we drew on Freirean theories of empowerment, Sen and
Nussbaums’s capabilities theory, and Bronfenbrenner’s ecological approach. As discussed elsewhere in this
chapter (see Box 7.2), new and unanticipated theoretical frameworks were found to be a closer fit with our
data—not unusual in qualitative research. We also focused on producing a preliminary grounded theory of
service engagement that was subsequently published (Padgett, Henwood, Abrams, & Davis, 2008).

The optimal means of theorizing involves making connections—across codes, themes, typologies, and
categories. Lists of themes are among the most common ways that qualitative findings are reported, but they
fall short if theorizing is the desired end. Charmaz (2014) proposes four criteria for appraising the value of a
grounded theory: theoretical plausibility, direction, centrality, and adequacy. Theoretical plausibility refers not to
the truth or accuracy of individual accounts but the extent to which they are plausible within a wider frame of
meaning. Direction is achieved when the conceptual framework starts to coalesce, and the conceptual building
blocks (or categories) show relationships. This requires some winnowing and distillation as less important
concepts or categories are cast aside. Centrality refers to this developing critical mass of interrelated ideas, and
adequacy means the developing theory has sufficient saturation. If not, theoretical sampling might be used to

269
flesh out the categories.

Since the majority of qualitative researchers do not produce new theory (nor do many aspire to or need to),
interpretation can lead to other valuable ends (see Figure 7.1 and the next section for an example). Theorizing
along the way is still advantageous to ensure that data are linked to broader concerns.

270
An Example of a Conceptual Schema
Figure 7.1 shows the conceptual model developed from Phase 1 of the NYSS. The model has a visual and
schematic logic, but narrative description is needed to make it come to life. Although space limitations
preclude full elaboration, certain features can be pointed out to illustrate thematic development. As shown,
the three themes are autonomy and choice (Theme 1), sanctuary and personhood (Theme 2), and personal
resources and capabilities (Theme 3). The subthemes elaborate on each theme and comprise its full content
(subthemes 1a and 1b were in vivo codes).

The dotted arrows show connections among the subthemes. The longest dotted arrow, for example, shows
that participants desire to “manage their own care” (subtheme 1b)—as in fiddling with medication dosages
and quitting drug abuse without treatment—was associated with reflections on advancing age and the need to
take charge of their lives (subtheme 3d).

The model’s theoretical uplinks are noteworthy. As described in Box 7.2, the code “living independently”
graduated to become a subtheme that aligned with the theory of ontological security (and expanded that
theory beyond its original scope). Themes 1 and 2 in the model are emic, that is, rooted in participants’
perspectives. Theme 3, on the other hand, is etic in that it captures our observations filtered through social
science theories and ideas. All of the subthemes under Theme 3 had enough empirical and conceptual salience
to support separate analyses and manuscripts for publication. These include subthemes 3a and 3c on gender
and trauma (Padgett, Hawkins, Abrams, & Davis, 2006), subtheme 3b on social network depletion (Hawkins
& Abrams, 2007), and subtheme 3d on the effects of aging and self-reflection (Shibusawa & Padgett, 2009).
Theories and concepts found applicable to the data and analyses included feminist, trauma, and social capital
theories as well as the life course perspective.

271
Salience Through Absence: What Is Not Said (or Seen)
One of the more intriguing paths to follow in interpreting qualitative data is to examine what participants do
not talk about (whether intentionally or unintentionally) and what is expected during observation yet remains
unseen or not experienced. Thorough knowledge of the subject matter and contextualization opens that door.
And it is a door worth opening.

Figure 7.1 Themes and Subthemes From Phase 1 of the NYSS

272
In the NYSS, participants were asked to talk about their lives and the challenges they had met. Given their

273
psychiatric diagnosis, the literature led us to expect that this would figure prominently in their narratives. For
the most part, however, this did not happen. Indeed, participants focused on their family troubles, health
problems, problems with the law, and struggles with substance abuse. Mental illness was talked about, but
most often in the context of unpleasant treatment experiences and medication side effects (Padgett, et al.,
2008).

Why is this salient? Characterizing persons by their diagnosis sets in motion expectations and causal
inferences. In the larger scheme of things, the incentivizing effects of large-scale funding for health-related
research in this United States channel researchers toward individual pathology rather than social or
environmental causes. In this context, mentally ill individuals whose life stories sparingly mention their
diagnosis can enlighten those of us assuming the psychiatric problem is paramount.

At times, a noteworthy absence is found in one source of data and not another, thus leading to questions of
which data are the most complete or trustworthy or whether both contribute different facets to a greater
whole. Consider a hypothetical ethnographic study of a gentrifying neighborhood in Harlem (in upper
Manhattan). Long a bastion of African American culture, Harlem has experienced recent in-flows of young
White families seeking affordable housing. Community leaders have condemned this “invasion” and asked for
help in ensuring that longstanding residents are not uprooted or made homeless by rising rents. Your field
observations in the neighborhood show a relatively smooth entrée of the White families with little hostility.
This absence of antipathy is given attention in your field notes. At the same time, in-depth interviews show
deep suspicions on both sides of the racial divide. A peaceful transition is happening but not without some
trepidation over gentrification.

A couple of caveats apply here. First, the difference between what is not said and not seen should be noted.
When participants are interviewed, they make decisions—consciously or unconsciously—about what to reveal
and what to leave out. Self-editing such as this gives the researcher a tangible parameter in which to work.
Relying on observation, one cannot avoid editing since we cannot see and record everything. Thus, what is
unseen could well be a product of the observer’s inattentiveness. Second, the ability to identify what is missing
(and salient) and what is present (and salient) depends on having broader knowledge of the phenomenon.
Calling attention to the unsaid or unseen has impact when that broader context is supportive.

274
Feminist and Other Critiques in Interpretation
The powerful undertow exerted by inequalities based on gender, race, ethnicity, sexual orientation, and
disability surely deserves researchers’ consideration—and it has received much attention in recent decades.
Fonow and Cook (1991) define feminist scholarship as incorporating reflexivity, social action, emotional
connectivity, and grounding in lived experience. Patty Lather, Shari Hesse-Biber, and other feminist
researchers caution against description-only research as falling short of their interpretive potential in exposing
the inequalities of patriarchy.

Similarly, critical race theory highlights the damaging effects of White privilege that inhere in laws,
institutions, and social structures (Delgado & Stefancic, 2011; Ladson-Billings, 2000). Queer theory has been
fertile ground for social and literary criticism of cisgender power, and qualitative methods have been adopted
to further these ends (Gamson, 2000). In The History of Sexuality, Foucault (1976), introduced the concept of
biopower, a term referring to the control of governments seeking to ensure a healthy workforce. Power over
bodies, according to Foucault, became a regulatory function that reinforced the state’s authority. Foucault’s
writing inspired broader indictments of post-1970s neo-liberalism typified by the Reagan and Thatcher
administrations in the United States and the United Kingdom where less government interference is desired in
unleashing economic and political power, but more government interference is desired in suppressing social
and sexual freedom.

Common to critical interpretation is the obligation to go beyond lived experience to bring in history and the
impact of oppressive institutions and structures. This is going not only up and out but back and forth, drawing
in information from varied sources. For qualitative researchers, this is a tall order, and the critical lens is
enhanced by use of documents as well as interviews and observation.

Sometimes, a critical perspective is awakened during analysis. While reading transcripts in the NYSS, we
noticed that African American men frequently referred to women as “females,” as in “I would like to have a
relationship with a female” or “Females have been a problem for me.” This distancing term of reference fit
with a growing awareness that a gender divide characterized participants’ life experiences. Summarized pithily
as “outlaws versus outcasts,” this divide cast homeless mentally ill women in the gendered role of victim as well
as perpetrator (the latter condemning them as bad mothers, sexually promiscuous, and unfeminine).
Compared to women’s status as “outcasts,” men could be “outlaws” (drug dealers, pimps, illegal vendors) in
masculine roles celebrated rather than denigrated in the wider culture (Padgett et al., 2006). Further evidence
of the gender divide was that homeless women’s access to the underground economy was severely constrained
(recycling cans or survival sex frequently the only options). Homeless men, like the other poor men in their
neighborhood, could sell drugs or contraband, become a street vendor, or work in off-the-books construction
jobs. This observation of “defeminized” women and still-masculinized men aligned with ethnographic
research on African American men and a “code of the street” (Anderson, 2000).

In narrative and discourse analyses, the interpretation phase may bring in master narratives and counter
narratives (Bamberg, 2004). Master narratives are dominant tropes such as the “American Dream” or

275
“America as a melting pot.” In the Southern United States, a master narrative among many White citizens
was that of Confederate Army resistance against the tyranny of the North rather than defending racism and
slavery in the South. Counter narratives would oppose such hegemonic discourses (Bamberg, 2004), opening
the door through interviews with elderly Southerners to document the toll of Jim Crow laws and other deeply
racist actions. Similarly, interviews with unemployed Americans frequently give the lie to the American
Dream master narrative (Pederson, 2013).

Observations of the counterintuitive and unexpected incite new ideas and fresh perspectives that can fuel
critically informed interpretations. Debunking, for example, is useful to push off from the side of the pool and
swim against the tide (forgive the mixing of metaphors). Qualitative methods are well suited to this as they
afford the flexibility, depth, and agility to recognize when the received wisdom or popular perceptions obscure
and mislead. Elliott Liebow’s Talley’s Corner (1967) became an enduring classic in large part because it
upended traditional notions of poor African American men as willfully unemployed.

The prominence of social justice values in critical interpretation is its signature contribution. Although this
brings the risk of overreach to serve ideological ends, critical perspectives fill an important gap in
interpretation. Critical analyses do not preclude using quantitative methods (statistics can be used to condemn
economic inequality in a way qualitative methods cannot match), but their close fit with qualitative methods is
indisputable. Indeed, the passion of many proponents of qualitative methods has come from this alignment.

276
Rhetorical Devices
Rhetorical devices can help formulate ideas during interpretation. Qualitative researchers are fond of using
metaphors and analogies (Lakoff & Johnson, 1980; Ryan & Bernard, 2003). The phrase “going AWOL”
illustrated our use of an analogy in the NYSS. It began as an in vivo code, its origins in military usage
(referring to being absent without leave) broadened to include a variety of conditions surrounding the sudden
departure from treatment programs. A resident may “go AWOL” from a rehab center, a residential program,
or a shelter—usually due to an unwillingness to abide by stringent rules set against the attraction of almost
anywhere else (including the streets). What lifts an analogy to the interpretive level is its capacity for larger
meaning. Thus, “going AWOL” could be construed more broadly as exercising agency in one’s life when the
options are few.

The participant can also use salient metaphors. In the Harlem Mammogram Study in which I was involved in
the late 1990s, some of the women with breast cancer used a metaphor to describe how cancerous tumors took
root and spread, one saying “I had a garden growing in there.” This evocation of a potentially deadly disease as
originating in “seeds” (cancer cells) that sprout and send tendrils into the body evokes organic yet uncontrolled
growth. The implications—had we been able to follow through with such a line of inquiry—might be that
chemotherapy, that is, “poisoning the weeds,” would be more acceptable than radiation with its reputation for
tissue damage.

Recalling the framing devices mentioned earlier in this chapter, turning points and epiphanies are popular
rhetorical devices around which to build interpretations. During an interview I conducted with a leading
homeless advocate from a Western state, he mentioned having “epiphanies” when describing his journey of
awareness that occurred during his expanding role in changing the state’s service system to be more responsive
to homeless persons’ preferences.

Interpretation, regardless of approach, requires building upon immersive interaction with the data. In this
way, dominant narratives, phenomenological essences, and deconstructed discourses join coding as the
foundation for abstraction and induction. As mentioned earlier, interpretation also draws in new ideas,
concepts, and theories. As illustrated in Box 7.2, referring back to the literature can reap rewards in linking
one’s analyses up and out.

277
Box 7.2 The Life of a Hard-Working Code: Linking “Up and Out” to
Theory
“Living independently” became a workhorse of a code in the NYSS, ultimately becoming a subtheme. Alternative code labels such as
“need for housing” or “desire for autonomy” might have sufficed, but they did not do justice to participants’ descriptions of what it
was like to have their own apartment—either in the present or in a hoped-for future.

Examining the many coded excerpts filed under “living independently” revealed more than just material comforts being invoked (or
sought). Women talked about no longer needing to trade sex for shelter and being able to avoid abusive male friends. Men spoke of
the freedom of having a key with a lock on the door and of having clean clothing every day. Men and women alike noted the relief of
not being under constant surveillance and supervision. These psychological benefits extended beyond the apartment walls, as
participants noted how gratifying it felt to go out and know they had a safe place to return where they could cook their meals, watch
television, entertain friends, or just be alone.

Glaser (2002) writes about the staying power of concepts and their capacity to link a study’s findings “up and out” to larger issues and
ideas. Attempting to better understand what “living independently” meant in the context of housing status, I searched the public
health and urban planning literature and came across the theory of ontological security. This produced a classic “aha” moment.

R. D. Laing (1965) wrote that ontological insecurity was a problem for persons with schizophrenia whose mental illness deprived
them of stable functioning and identity development. Sociologist Anthony Giddens (1990) developed a broad theory of ontological
security for the postmodern era, arguing that many persons need (but lack) the constancy in their social and material environment
that engenders self-actualization and faith in one’s future. Interestingly, the literature revealed that empirical research on ontological
security was largely confined to studies of home ownership (versus renting) in New Zealand and the United Kingdom. In this
context, the leap from sleeping on a park bench to one’s own apartment seemed especially momentous. Dupuis and Thorns’s (1998)
markers of ontological security related to having a home provided sensitizing concepts for me to use as I returned to the NYSS Phase
1 data to recode and reanalyze. These markers included constancy, day-to-day routines, a sense of control due to a lack of
surveillance, and a secure base from which identities can be constructed (Dupuis & Thorns, 1998).

The markers mapped onto the data beautifully. Moreover, as often happens, a new code emerged inductively from the analysis.
Labeled “what’s next,” this code captured the anxiety felt by participants who had left behind the urgency of survival to face an
uncertain future amidst the cumulative disability brought on by poverty, mental illness, and social exclusion (Padgett, 2007).

Commentary: This story of a code that proved to have interpretive traction exemplifies how outside theories may be brought into the
study to help illuminate and refract meaning from the data. In this instance, it also shows how the outside theory can itself benefit
from the association and gain broader explanatory power. Although most codes do not have this salience, vigilance is needed to
discriminate the leaders from the rank-and-file.

278
Negative Case Analysis and Disconfirmation
For some qualitative studies, coding and analyses set the stage for theorizing and developing hypotheses to
assist in interpretation. This, in turn, introduces negative case analysis as a means of testing provisional
hunches or ideas. The search for negative cases roughly corresponds to the quantitative researcher’s reliance on
a null hypothesis, that is, searching for falsifying evidence tests emerging theories. In this way, we become our
own devil’s advocate. This follows the logic of philosopher Karl Popper that hypotheses are not truly verified
but only supported in the absence of refuting evidence. As Popper noted, even if we see a million white swans,
the sight of only one black swan will disconfirm our “white swan” thesis. Similarly, Albert Einstein said, “no
amount of evidence can prove me right, and any amount of evidence can prove me wrong” (quoted in Miles &
Huberman, 1994, p. 242). Closely linked to grounded theory, disconfirmation tightens and extends the
analyses.

Grounded theorists turn to theoretical sampling to search for negative cases. Theoretical sampling involves the
careful selection of cases or settings that will help test emergent hypotheses. Perhaps you are studying
adolescents who drop out of high school and notice a pattern (e.g., that boys say their reason for leaving was
bullying and intimidation, and girls say their family responsibilities interfered with school). Further analyses
lead to developing a provisional theory of gender differences in “push versus pull” factors, and you wish to
refine it further. This could entail seeking out teens who stayed in school to ascertain if gender differences in
these stressors hold and if their intensity is less (hence they had stayed in school).

Theoretical sampling can also involve seeking new types of data and reinterviewing study participants with
new questions. Finding a disconfirming case or other anomalous information need not lead to completely
discarding the theory, but it should prompt serious reconsideration. If the countervailing evidence starts to
pile up, it may be time to let the provisional theory go.

Of course, negative case analysis should not occur in a vacuum—extant research may bolster or undermine a
budding hypothesis. If a lot is already known on the topic, does this mean negative case analysis is unnecessary
and one need only consult the literature? Not at all. In the first place, anomalous cases within one’s data
should be given their due. In qualitative research, it is not about quantity of evidence but quality. Secondly,
disconfirmation can set into motion a more focused and deeper inquiry into the contexts of the observed
associations (or discrepancies).

It is important to distinguish between disconfirming evidence (cases that refute an emerging theory) and
discrepant evidence (cases that refine an emerging theory; Goetz & LeCompte, 1984). In the first instance, the
exceptional case disallows the rule (e.g., the black swan). In the second instance, it proves the rule but refines
and expands it. The line between disconfirming and discrepant cases is blurry at times (Ely, Anzul, Friedman,
Garner, & Steinmetz, 1991). Theories may become so refined and spread so thin that their explanatory value
begins to sag under the weight of the evidence. A researcher may also be so taken with negative case analysis
that she disconfirms every theme and achieves the dreaded state of analytic paralysis. Somewhere there is a
happy medium—a mixture of enthusiasm tempered by skepticism. In the meantime, it is probably better to

279
err on the side of caution than to throw it to the wind.

Where does this leave qualitative researchers when questions of causation arise? Experts run the gamut of
opinions about this, ranging from relative enthusiasm (Miles & Huberman, 1994) to serious doubt (Lofland
& Lofland, 1995). As mentioned earlier in this chapter, causal explanations are not typically included in the
promise of qualitative inquiry. On the sidelines, anti-positivist skeptics question whether the search for
causation is plausible or desirable, given the postmodern premise that facts are “fictitious” (Lofland &
Lofland, 1995). As with so many contested issues in qualitative research, one’s position depends in large part
on one’s epistemological stance. My inclination is to be cautious and not make promises qualitative research
cannot (and probably should not try to) keep.

280
The Six Approaches Revisited
In what is by now a familiar mantra, we begin with noting the importance of paradigmatic stance as an
influence on how the study enters its interpretive stage. For some paradigmatically passionate researchers, this
may supersede the qualitative approach being used. Constructivists, for example, view interpretation as more
or less synonymous with data analysis and the researcher as an intrinsic part of the story regardless of
approach.

Interpretation is one of the ways that a qualitative study is distinct from journalistic description. Narrative or
discursive findings often involve “zoom in–zoom out” techniques in which meticulous dissection of a discourse
or conversation is situated within larger social and cultural contexts. The interpretive act of “zooming out” is a
necessary corrective to the hermetic immersion required of “zooming in.”

For their part, phenomenological interpretations identify essences and interpret the lived experience based
upon intensive immersion. The case study researcher and ethnographer share a holistic stance to interpreting
data. As discussed in Chapter 2, intrinsic case studies are more likely to stay at the descriptive level while
instrumental case studies serve as touchstones for larger social, political, or economic insights. Similarly, the
graphic description of ethnography is essential but has larger meaning only when contextualized theoretically.
In contrast, action-oriented researchers focus their interpretations on concrete goals and are thus more likely
to favor communal problem resolution over interpretive latitude.

The postpositivist paradigm associated with traditional grounded theory offers the most explicit steps for
transitioning from early to later stages of analysis ending in interpretation. Charmaz (2014) refers to focused
coding as the time when open codes are winnowed down. Focused codes tend to be more conceptual and less
tied to individual responses. They may result from aggregating two or more open codes under a single label
but just as often emerge from a single open code with high salience. For example, Charmaz traces the origins
of her most salient category (“identifying moment”) to its initial status as a code label for when a person with
chronic physical illness is reminded by others of his disability. Originally having only a negative valence, the
code was broadened to include positive experiences and was ultimately promoted to category status. Although
most codes serve out their duty in the descriptive trenches of analysis, an occasional one will make the jump to
higher levels.

In grounded theory, axial coding specifies the properties of the category and its subcategories, and selective
coding entails searching for relationships between categories as the final step toward completion of a grounded
theory (Strauss & Corbin, 1994). Just as open coding parses data into digestible bits, focused and axial coding
start the process of reintegration by creating a preliminary conceptual framework. In practice, few qualitative
studies use axial coding because it is demanding and, according to some, overly prescriptive (Kelle &
Erzberger, 2004). The same applies to selective coding, a process of selecting and refining “core categories”
alone and in relationship to one another (Strauss & Corbin, 1994)—although this is desirable when pursuing
a fully grounded theory as the study’s endpoint.

281
In her study of disordered eating among young women, Hesse-Biber (2004) noted how her codes went from
the literal, for example, “boys will love me,” “thinnest and prettiest” to analytic focused codes, for example,
“thin rationale” (p. 412). The latter encompassed several ideas including beliefs that thinness was healthier,
more likely to produce wealthy husbands, and provoke the admiration of other women. Hesse-Biber also
ventured “outside” of the data to take note of a growing media preoccupation with thinness and how ultra-
thin body images of models filtered into the young women’s lives.

Clustering codes can be done as part of an emerging theory or to juxtapose against a priori theoretical
frameworks (Boyatzis, 1998). Clustering usually reflects shared underlying conceptual content, but it may also
be used to show temporal or quasi-causal relationships. For example, the codes “hitting bottom,” “going cold
turkey,” and “maintaining sobriety” could be clustered together as “overcoming addiction.” This sorting not
only raises the level of abstraction but also helps to winnow out extraneous material. It also starts the process
of hierarchy building if larger themes are desirable. The ultimate goal is parsimony in one’s findings, so
condensation is normal. However, the caveat applies here once again that one should not force the data.

Researchers adopting any of the qualitative approaches often turn to schemas and diagrams to display categories
and relationships (Miles & Huberman, 1994). Arraying key concepts as nodes awaiting connections can
greatly enhance analysis through showing interrelationships and hierarchies of meaning. Directional arrows
connect the nodes, and a pyramid or circular arrangement can give order and precedence.

Strauss and Corbin (1994) offer a template schema for developing a grounded theory that includes identifying
conditions, actions, and consequences. As an example, this schema could be applied to an understanding of why
homeless mentally ill clients “go AWOL” from residential programs, including the conditions or
circumstances surrounding this act (rule-breaking, relapsing into drug abuse, etc.), the specifics of carrying it
out (when, where, with whom, etc.) and the consequences that ensue (e.g., return to the streets or shelters).
Oktay (2012) suggests constructing a pyramid-type hierarchy displaying subcodes at the bottom level, then
codes, then code clusters or focused codes, and then categories or themes at the top.

Patton (2002) advises ensuring that the emergent themes have internal homogeneity (data indexed by them
dovetail closely) and clear boundaries (minimal or no overlap across categories). Ideally, the themes (and
subthemes) capture all salient information, that is, they become saturated. Although rarely done, thematic
development may employ quantification as part of the process. Box 7.3 offers one such example.

282
 Box 7.3 An Example of Quantified Thematic Development
Barkin, Ryan, and Gelberg (1999) describe a deductive-inductive approach to analyzing interview data from pediatricians,
community leaders, and parents in South Central Los Angeles inquiring about what doctors can do to prevent youth violence. Given
the narrow scope and limited time for the study, the researchers started out with three a priori themes: potential (for doctors to
address youth violence), barriers (to addressing youth violence), and resources (for addressing youth violence). These themes were
used by two team members to identify statements associated with each (numbering 84, 74, and 41, respectively) made by the 26
study participants. For subtheme identification, this deductive approach shifted to inductive as four coders separately pile-sorted the
three sets of thematic statements to come up with subthemes. Three of the four coders were naive (blinded) to the study’s goals.
Each statement was scaled 0 to 4 depending on the number of coders who placed it in a subtheme pile. The researchers then turned
to multidimensional scaling and cluster analysis to identify the final set of subthemes. The distributions of these subthemes were
displayed by interview group to show how frequently pediatricians, community leaders, and parents endorsed them. Although the
authors did not use the term, the early phases of data analysis resembled content analysis insofar as they were identifying previously
selected phenomena and counting their frequency.

Others prefer to display themes and subthemes as nodes and networks with linkages in between. (See Figure
7.1 for an example.) Researchers who wish to go the distance and develop a grounded theory are obliged to
pursue the more demanding steps of testing their ideas and ensuring that all connections cohere into an
integrated model.

Among the six approaches, grounded theory clearly has the most developed procedures for moving from
coding analyses to interpretation. This is not to say that the other approaches are not procedural (or that
grounded theory is rigidly prescriptive), but they are less codified. Across all of the approaches, interpretation
requires elevating one’s sights, looking up and around while staying grounded in the data and analyses that
came before.

If there is any common element to interpretation across the qualitative approaches, it is probably thematic
analysis. Pattern recognition is part of human cognition. It facilitates abstraction and conceptualization.
Patterns coalesce into themes as the analyses move toward closure. For qualitative researchers intent on
higher-order abstraction and theorizing, presenting a list of themes is like erecting pillars without completing
the structure they were designed to support. At the same time, that overarching structure needs to have
meaning and purpose. Otherwise, the pillars alone may be enough.

283
Qualitative Meta-Synthesis as Meta-Interpretation
As the number of qualitative studies has grown exponentially in recent years, attempts to aggregate them
through meta-syntheses have become increasingly common beginning with seminal work by Noblit and Hare
(1988) and continuing with more recent proponents (Erwin, Brotherson & Summers, 2011; Major & Savin-
Baden, 2010; Sandelowski, Docherty, & Emden, 1997; Thorne, Jensen, Kearny, Noblit, & Sandelowski,
2004; Walsh & Downe, 2005). Similar to meta-synthesis and meta-analysis is the systematic review, the latter
explicit in the selection process but less dependent on standardization in aggregating the findings once the
studies have been identified.

The move to synthesizing findings to promote evidence-based practices signified by the Cochrane and
Campbell Collaborations (http://www.cochrane.org and http://www.campbellcollaboration.org) has gained
momentum, with summative reviews ranging from interventions for survivors of sexual abuse to social skills
groups for persons with autism to the latest in cardiac rehabilitation. Efforts to bring order to this burgeoning
“industry” of meta-reviews have led to criteria such as Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA), an evidence-based set of items recommended for reporting in systematic reviews
and meta-analyses (http://www.prisma-statement.org). Relatedly, the Critical Appraisal Skills Program
(CASP) was developed as an online resource for persons seeking the latest information on medical and other
treatments (http://www.casp-uk.net). CASP includes qualitative studies along with case control,
experimental, and other quantitative designs. Like their quantitative counterparts, qualitative meta-syntheses
have been concentrated in health research, but they are not restrictive in focus.

The rationale for meta-syntheses is simple: Without some means of connecting the dots, qualitative findings
remain atomistic and unrepresented in broader representations of a phenomenon (Erwin et al., 2011). Beyond
describing commonalities, meta-syntheses help shine the light on gaps in understanding and point the way to
future research (Major & Savin-Baden, 2010; Zimmer, 2006). Unlike meta-analyses, meta-syntheses are
subject to varying opinions about their utility and methods. This is hardly surprising since many skeptics view
qualitative studies “as resistant to synthesis as poems” (Sandelowski & Barroso, 2003, p. 7).

Nevertheless, methods of qualitative meta-synthesis have been elucidated, beginning with a clear definition of
the phenomenon of interest and the scope of the review. Online searching can locate virtually all published
and many unpublished works on a particular subject (a knowledgeable librarian can be invaluable). Databases
such as Google Scholar, Embase, Web of Science, PsyINFO, PubMed, and Dissertation Abstracts are
explored using key word search terms with Boolean-type and or refinements.

The balancing act involves casting a net that is wide but not too wide. Even a narrow topic of interest can
generate many articles, book chapters, theses, and reports (including gray literature) that have to be screened
according to the selected inclusion/exclusion criteria. In her synthesis of women’s experiences of treatment for
perinatal substance abuse, Hines (2011) retrieved 33 studies and excluded three because they lacked direct
excerpts or quotes—a key requirement in keeping with the Noblit and Hare approach (1988). As described in
Box 7.4, the numbers of retrieved articles can be in the hundreds.

284
After the netted catch is winnowed down, a common next step is to evaluate the studies methodologically.
While appraising quality is a contested arena in qualitative research, to ignore this step is to treat all studies
equally, thereby lending importance to weak studies and downplaying the importance of rigorous ones.
Checklists of criteria come into play here with each study assigned a score based upon its degree of rigor and
transparency (Erwin et al., 2011). Some meta-syntheses set a cut point such that very weak studies are
excluded; this should be decided in advance and made part of the protocol.

The next step is to synthesize the studies (that are by this time considerably smaller in number). Noblit and
Hare’s (1988) original method entailed identifying common phrases, themes, concepts, or metaphors, but
later iterations include constant comparative analysis and multiple case study analysis (Erwin et al., 2011). In a
meta-synthesis of the experiences of parents of children treated for kidney disease, Tong and colleagues
distinguished between first-, second-, and third-order constructs in their interpretation of the studies. First-
order constructs were drawn from the quotes of the original participants, second-order constructs were the
themes reported by the original researchers, and third-order constructs were the meta-themes developed by
the synthesizing researchers (Tong, Lowe, Sainsbury, & Craig, 2008). As with primary data analysis, meta-
syntheses may use QDA software for coding articles in a way similar to coding of transcripts.

Qualitative meta-syntheses use a variety of analytic strategies, and the proliferation of terms has been
considerable. In their “meta-study” of qualitative meta-syntheses, Barnett-Page and Thomas (2009) located
203 articles using nine different approaches: meta-narrative synthesis, critical interpretive synthesis, meta-study,
meta-ethnography, grounded formal theory, thematic synthesis, textual narrative synthesis, framework synthesis, and
ecological triangulation. The authors noted that several of the approaches used overlapping techniques (not a
surprise). Common to all is the use of cross-case interpretation (Walsh and Downe, 2005).

285
 Box 7.4 Family Experiences of Child Death in a Pediatric Intensive
Care Unit (PICU): A Meta-Synthesis
The untimely death of a child in a pediatric intensive care unit (PICU) brings intense suffering enacted in a sterile high-technology
environment. To examine this phenomenon, Butler and colleagues (2015) conducted a qualitative meta-synthesis that was registered
with the International Prospective Register of Systematic Reviews, also known as PROSPERO
(http://www.crd.york.ac.uk/PROSPERO). Showing their allegiance to the biomedical format common to many such reviews, Butler
et al. adopted the guidelines of PRISMA (http://www.prisma-statement.org) (Butler, Hall, Willetts, & Copnell, 2015).

Their review began with consulting an expert medical librarian before searching five databases for the years 1990 to 2014: CINAHL
Plus, OVID Medline, Scopus, PyschINFO, and Embase. This yielded 962 articles. The reference lists of retrieved articles were also
searched (6 more articles found in this way) as was gray or unpublished literature (16 more articles). Reviews of the abstracts of the
984 articles showed 328 duplicates and 548 criteria-based exclusions, leaving 108 for full-article scrutiny. Of these, only 15 remained
in the final synthesis group (the others were not in English, not qualitative, lacked ethics committee approval, were incomplete, etc.).

To ascertain methodological quality, the authors used the CASP checklist. The final group of 15 studies was thematically analyzed
by four researchers using a method devised by Thomas and Harden (2008) that is similar to grounded theory coding. The primary
theme “reclaiming parenthood” and its requisite subthemes of “being supported,” “parenting after death,” and “being a parent in
PICU” pivoted around the need to make meaning of the child’s death when medical staff were often focusing on the instruments
and equipment of the PICU.

Commentary: This study exemplifies the dramatic filtering needed to identify eligible studies before the synthesizing begins (984
articles winnowed down to 15). Though more flexible (and more forgiving of diverse research designs), the protocols associated with
qualitative meta-syntheses might strike some qualitative researchers as mechanized. The codified procedures of meta-syntheses do
indeed betray their biomedical origins. But once one leaves the discretionary zone of a traditional literature review, some degree of
standardization is necessary.

The complications of conducting meta-syntheses are not easily ignored when it comes to decisions around
screening and analytic methods. Should one include mixed method studies? How far back in time should one
go? (Journals began online referencing in the early 1990s.) How does one distinguish multiple qualitative
reports from a single parent study? What degree of methodological weakness should be tolerated in the spirit
of inclusiveness? Given the array of analytic options, which are best suited to the task? How does one choose
among these options when there are so many similarities?

And there are larger concerns. Problems of representation in meta-synthesis can arise from conveying a
message of certainty and generalizability that the component studies were not intended to convey in the first
place (Thorne, 1998). For example, most qualitative studies of persons with schizophrenia have relied heavily
or exclusively on White respondents even though African Americans are disproportionately diagnosed with
the illness. Synthesizing findings from these samples will almost certainly misrepresent the experience of (and
treatments for) schizophrenia.

Even the most avid qualitative synthesizer knows that the costs of standardization are substantial—losing the
richness and nuance of qualitative studies can stifle their impact. Yet the meta movement is a train that has
left the station, and avoiding that journey risks sidelining qualitative studies from the important issues of the
day.

286
287
In Pursuit of Methodological Transparency
As will be discussed further in Chapter 9, qualitative reports often obscure full description of their methods.
This state of affairs reflects a practical problem (word and space limits) as well as a tradition of eliding
methodological description in the absence of agreed-on procedures. Not surprisingly, the contestation over
qualitative methods has also had an impact, with postmodern critics deploring “methodolatry” (Janesick,
2000) and pragmatists urging greater transparency. Concerned about this trend, British ethnographer Paul
Atkinson (2005) pointed to an American fixation with rejecting methods that “smack of formal analysis,”
substituting in their place “vague ideas of experience, evocation and personal engagement” (p. 4). This leaves
novice researchers struggling to emulate their role models from afar with inadequate instruction.

This chapter and the previous chapter represent attempts to enhance transparency in the analysis and
interpretation of qualitative data, with the admonition that one is often reporting post-hoc as well as
proposing in advance. Demystifying qualitative methods makes them more accessible and facilitates cross-
disciplinary communication (Ryan & Bernard, 2003). To be sure, an overemphasis on documentation of
methods can lead to superficiality and bland findings. Yet a study’s conclusions inspire much greater
confidence when built on methodological transparency.

288
Summary and Concluding Thoughts
Interpretation gains momentum as the analysis proceeds and connections are made; memos log this process
and propel it forward. The relationships may be temporal (as in stages and phases) or relatively static
conceptual networks arrayed in horizontal or vertical (hierarchical) fashion. Searching for disconfirming cases,
debunking, and metaphors helps open up the process and keep it fresh. Interpretation brings in the larger
context (i.e., applying interpretive frames from the literature and from the realms of practice and policy).

This chapter’s focus on interpretation belies the iterative nature of data collection, analysis, and write-up.
Early stages of data analysis give rise to theorizing and this fuels the interpretations that segue into the final
write-up (writing is itself a form of interpretation). Sometimes we go back to data collection to extend and
tighten the interpretations being made. While the feedback loops may seem endless at times, closure does
come, as it must. Ideas and themes—even theories—start to gel; extraneous material is set aside and findings
take shape. Interpretation brings salience to the surface and showcases it.

289
Exercises
1. Consult a qualitative journal listed at the end of Chapter 1, and locate an article that reports on a
grounded theory study. How did the authors proceed from coding to interpreting their findings?
2. With reference to the framing devices linked to social work concerns listed in this chapter, how might
social justice values inform a study of homelessness? A study of depression among single mothers? A
study of transgender runaway youths?
3. Examine Figure 7.1, and describe how a theory might be developed from the interrelationships shown
between the themes.
4. Consider a social or health problem of interest to you, and go to either the Cochrane or Campbell
collaboration websites to find out if that problem has been meta-analyzed (or meta-synthesized). If so,
what are the findings?

290
Additional Readings
Becker, H. (1998). Tricks of the trade: How to think about your research while you’re doing it. Chicago, IL:
University of Chicago Press.

Bernard, H. R., & Ryan, G. (2009). Analyzing qualitative data: Systematic approaches. Thousand Oaks, CA:
Sage.

Charmaz, C. (2014). Constructing grounded theory: A practical guide through qualitative analysis (2nd ed.).
Thousand Oaks, CA: Sage.

Flick, U. (Ed.). (2013). The Sage handbook of qualitative data analysis. Thousand Oaks, CA: Sage.

Lofland, J., & Snow, D. A. (2005). Analyzing social settings: A guide to qualitative observation and analysis.
Belmont, CA: Wadsworth.

Silverman, D. (2015). Interpreting qualitative data (5th. ed.). Thousand Oaks, CA: Sage.

291
8 Strategies for Rigor and Trustworthiness

292
Evaluating Qualitative Research
Consensus around what is high quality qualitative research has been elusive. To constructivists, a consensual
set of criteria raises the specter of intrusive oversight and enforced conformity. To pragmatists, standards of
quality are needed, but formulating them is an admitted work in progress. To critics of qualitative methods, a
lack of uniform standards means that findings are little more than anecdotal evidence (Adams, 2013). A
critical distinction arises at this point: Evidence of quality emerging from a completed study is not the same as
building into a study’s design deliberate efforts to improve quality. Evaluative standards are applied to
completed studies; strategies for rigor are pursued during the study.

This chapter is devoted to the dynamic discourses surrounding standards of quality in qualitative methods. At
one end of the epistemological continuum is an interpretive, arts-and-humanities approach that does not
subscribe to standards beyond broad aspirational goals. The opposite end of the continuum—a uniform quasi-
prescriptive approach—has been bolstered in recent years by the development of evaluative checklists
prompted by the rise of systematic reviews and syntheses. In between are big tent approaches with more
inclusive parameters (Tracy, 2010). Social work researchers can be found all along this continuum with
requisite attention to social work values (Drisko, 1997; Gilgun, 1994; Seale, 2002).

293
Evaluative Criteria and External Standards
What Schwandt (1994) derisively referred to as “criteriology” or the too-eager pursuit of standards has become
the focal point for lively debate. Lincoln and Guba (1985) were in the vanguard in developing separate criteria
applicable to qualitative methods (which they initially called naturalistic inquiry). Drawing parallels with
quantitative research, they proposed, credibility, transferability, dependability, and confirmability as alternatives
to internal validity, external validity, reliability, and objectivity. Together, these connoted the trustworthiness of
a qualitative study.

Credibility is the degree of fit between respondents’ views and the researcher’s description and interpretations.
Transferability refers to generalizability, not of the sample (as in quantitative terms) but of the study’s findings.
External validity is not a priority because the focus is on subjective meanings and depth over breadth
(Donmoyer, 1990). Auditability (or dependability) means that the study’s procedures are documented and
traceable—they need not lead to the same conclusions but should have a logic that makes sense to others.
Confirmability is achieved by demonstrating that the study’s findings were not imagined or concocted but,
rather, firmly linked to the data. Guba and Lincoln’s (1985) criteria did not receive universal acclaim or
adoption, but they served an important purpose in offering alternatives to replace ill-fitting quantitative
standards.

There are also standards closely aligned to what qualitative methods do best. Thus, qualitative studies are
much more likely to have ecological validity, that is, taking place in naturalistic conditions far from the
artificiality imposed by an experiment (Shadish, Cook, & Campbell, 2002). Another standard is evidentiary
adequacy (Erickson, 1986; Morrow & Smith, 1995). Defined as sufficient time spent in the field as well as the
density and diversity of evidence brought to bear, evidentiary adequacy can be a powerful contributor to a
qualitative study’s credibility.

294
In the Eye of the Beholder?
An appraisal of quality implies having a sense of what is construed as good and bad. Attention to
methodological rigor is balanced with attention to insightful interpretations—high quality presumes both exist
in equipoise. With respect to the importance of rigor in research, consider the distinction between the arts and
sciences. Art critics are the guardians of esthetic excellence, their taste and judgment intended to keep out the
less worthy (Becker, 1996). For museum goers, gallery visitors, and art buyers, beauty may be in the eye of the
beholder, but the beholder’s access to the art and its market value are largely determined by a small number of
judges and critics.

This art world example is presented to illustrate an esthetic and elitist means of determining quality. Such a
system of quality determination exists in contrast to what takes place in research where peer review is the
norm. Intended to reduce “art world” judgments, peer reviewers are typically blinded to the authorship of the
work and are expected to observe extant standards of quality such that arbitrariness and biases are reined in.
Without peer review approval and publication in a respected outlet, a study is consigned to the gray literature
at best, not the most prestigious way to gain respect for one’s work. Though far from perfect, peer review is
preferable to leaving quality decisions to a small elite group.

Qualitative researchers pay a price for this, as their work often falls victim to the lack of consensus on
standards. Running through the commentaries on rigor in qualitative research is a thread of doubt about
whether it is possible (much less desirable) to specify uniform standards that could be applied across the vast
landscape of qualitative methods. In a de facto way, qualitative researchers have been self-evaluating for a long
time without resorting to uniformity and conformity. Some studies make it into journal or book form, and
others are rejected. Who and what determine this?

The anti-authoritarian sentiment that pervades qualitative inquiry runs counter to art world elitism, but such
judgments are a likely outcome if consensual standards of quality are rejected altogether. When peers lack a
shared understanding or quality standard, opinions become uneven and arbitrary as reviewers fall back on their
own experiences and views of what is quality work. Of course, going too far down the road of uniformity runs
the risk of eroding what qualitative methods do best and the methodological pluralism that is their hallmark.

295
Attending to Rigor and Trustworthiness
Dictionary definitions of rigor bring to mind strictness and rigidity (as in rigor mortis). Indeed, the term rigor
has been criticized as exclusionary and rule-driven and thus incompatible with the thick description needed to
capture the human condition (Gergen, 2014). A leading qualitative social work researcher asserts that the
pursuit of rigor is a detour from social work’s intellectual history rooted in the work of Jane Addams, John
Dewey, and the Chicago School (Jane Gilgun, personal communication).

It is often easier to talk about what the pursuit of rigor is not rather than what it is. Because qualitative
research is not devoted to capturing a fixed reality, internal validity is an inappropriate criterion of rigor. Nor
are reliability and replication suitable criteria because they imply fidelity and repetitiveness rather than
flexibility. Replication (the capacity to produce the same results using the same methods) is valued in science,
but its premise is inconceivable in qualitative inquiry given assumptions of ever-changing contexts and
conditions.

Guba and Lincoln’s concept of trustworthiness (1989) comes closest to capturing this phenomenon of
accountability in qualitative research. A trustworthy study is one that is carried out fairly and ethically and
whose findings represent as closely as possible the experiences of the respondents (Steinmetz, 1991).
Trustworthiness is not a matter of blind faith; it must be demonstrated. Once again, the devil is in the details.

296
The Meaning(s) of Generalizability in Qualitative Methods
Generalizability is not a priority in qualitative studies, at least as it is understood in quantitative research
(Donmoyer, 1990). An emphasis on generalizing strips away the context that imbues a qualitative study with
ecological validity. Few would argue that Bronislaw Malinowski’s (1922) classic studies of the Trobriand
Islanders are deficient because they lack generalizability. The need for external validity is also called into
question in much the same way.

Maxwell (2002) sought to redefine generalizability as having different levels of meaning, only one of which
refers to extending one’s findings from the sample to a larger study population (the traditional basis of
statistical inference dependent on random sampling). Instead, he referred to widening circles of extrapolation.
For example, can the findings from a study of youth gangs in Los Angeles be applied to local gang members
who were not interviewed? What about youth gangs in Chicago or gangs in general? It is reasonable to inquire
about these outwardly radiating circles of inference, but few researchers—quantitative or qualitative—address
them.

Qualitative studies are more likely to be judged by the transferability of their findings to other settings and
contexts rather than to the larger population from which the sample was drawn. A qualitative study infers out
rather than up. By this I refer to the ways that qualitative studies have broader meaning and resonance
through their conceptual and theoretical contributions. Many populations of interest are not “parameterized”
such that a credible random sample can be drawn—and this is the least desirable option in qualitative research
in any event. This means that the study’s setting is described such that readers know in full about the contexts
of the study and make their own judgments about its value.

Transferability is construed in different ways within qualitative inquiry. According to Tracy (2010), it is when
readers can personally relate to the study’s findings and see parallels to their own experiences. Hochschild’s
(1983) classic study of “emotion work” in the service sector (flight attendants, store clerks, waiters, etc.)
appealed to many Americans familiar with these kinds of jobs where one has to maintain an emotionally even
keel while dealing with difficult customers and labor conditions.

Transferability also refers to impersonal extrapolations that are compelling due to the ecological validity of
what was reported. One need not have familiarity with a juvenile justice facility to relate to Abrams and
Anderson-Nathe’s (2013) portrayal described in Chapter 2. Qualitative studies provide the reader with a sense
of having been there vicariously. Beyond that, they contribute concepts and theories with resonance, for
example, Hochschild’s (1983) “emotion work,” Goffman’s (1961) “illness careers,” Anderson’s (2000) “code of
the street,” or Biehl’s (2013) “zones of social abandonment.”

297
Threats to the Trustworthiness of Qualitative Studies
Threats to trustworthiness in qualitative research fall under three broad headings: reactivity, researcher biases,
and respondent biases. Reactivity refers to the potentially distorting effects of the researcher’s presence on
participants’ beliefs and behaviors. Quantitative research uses distance and controlled conditions to protect
against reactive effects, but the intensity and closeness of qualitative research relationships make this a
constant concern. Reactivity is a direct threat to ecological validity (Hammersley, 2008).

Researcher biases emerge when observations and interpretations are clouded by preconceptions and personal
opinions of the researcher-as-instrument. Thus, investigators may choose informants who are simpatico with
their world view, may ask leading questions to get the answers they want, or may ignore data that do not
support their conclusions. Emotional pitfalls can also contribute to researcher biases. The unwary qualitative
researcher may veer too far in either direction—overly familiar or alienated—and thus lose effectiveness.
Beyond interrogating one’s self about biases and predilections at the outset of a study, reflexivity comes into
play in subsequent stages of a study.

Finally, there is the threat of respondent bias. This issue is a bit trickier to address because it implies that the
respondents’ truthfulness and candor are questioned (see Chapter 5 for an extended discussion of this topic).
Respondents may withhold information and lie to protect their privacy or to avoid revealing unpleasant truths;
a sensitive topic can prompt a desire to conceal or mislead. Alternatively, some try to be helpful and offer
answers that they believe we (and/or the larger society) want to hear. Interviewers may collude with this to
avoid giving offense or simply because some forms of “glossing” are easier to take in (Yanos & Hopper, 2008).
Rather than deliberately mislead, respondents may have faulty recall or interpret events in a way that conflicts
with what the researcher knows from another source. What does one do, for example, if the agency records
say a respondent relapsed and was in detox for a week, and he does not mention this when asked about his
recovery? Getting at the truth may be worth risking a confrontation, but it is usually better left alone, at least
until a more propitious time when rapport and trust have deepened. Focusing too literally on truthfulness can
turn the researcher into an interrogator.

These three types of threats to trustworthiness—reactivity, researcher bias, and respondent bias—affect all
studies, whether quantitative or qualitative. The intensity of the research relationship and the pivotal role of
the researcher-as-instrument place qualitative studies in greater jeopardy with regard to the first two.
Fortunately, there are a number of ways to build confidence in the study’s methods.

298
Strategies for Rigor in Qualitative Research
Even when viewed favorably, qualitative methods are frequently misunderstood. I have encountered my share
of “friendly doubts” voiced by colleagues who wonder about small sample sizes and flexible methods. One way
to help alleviate doubts is to incorporate a section into the proposal explaining the rationale for using
qualitative methods (Marshall & Rossman, 2010; Morse, 1994; Munhall, 1994). If externally imposed, this is
an unfair burden (quantitative researchers need not do this). But when space permits, such an explanation is
an opportunity to educate the audience and convey a sense of mastery.

Describing a rationale for one’s choice of methods is not the same as taking specific actions over the course of
the study. In the following paragraphs, I will discuss six strategies for enhancing rigor and trustworthiness in
qualitative research. Although not all are relevant or feasible for any given study, they represent an array of
techniques consonant with the aims of qualitative research. Culled from the literature on qualitative methods,
each strategy addresses one or more of the threats to trustworthiness described earlier.

299
Prolonged Engagement
Arising from the early days of anthropological fieldwork, prolonged engagement has come to be a defining
characteristic of qualitative studies regardless of where they take place. As shown in Figure 8.1, prolonged
engagement helps to ameliorate reactivity and respondent bias. Thus, the effects of the researcher’s presence
dissipate considerably when she spends long periods of time in the field and becomes accepted (or at least
tolerated).

Prolonged engagement makes withholding information or lying by respondents less likely. As Eliot Liebow
(1993) noted in his study of homeless women, “lies do not really hold up well over long periods of time” (p.
321). A trusting relationship between researcher and respondents reduces the motivation as well as the
opportunity for deception. An experienced researcher can often tell when respondents are shading the truth or
lying, but it usually takes more than one encounter to do so. For interview-based studies, prolonged
engagement may not be possible (as discussed in Chapter 5, conducting more than one interview is a step in
the right direction).

One drawback of prolonged engagement is the risk of researcher bias. Researchers can go too far in either
direction—they can “go native” and lose all interpretive distance or experience the “familiarity breeds
contempt” problem. Nevertheless, when properly reflexive about these issues, the advantages far outweigh the
disadvantages.

300
Triangulation of Data
The term triangulation, borrowed from navigational science and land surveying, originally referred to using
two or more sources to achieve a comprehensive picture of a fixed point of reference. Four types of
triangulation were outlined by Denzin (1978):

1. Theory triangulation: The use of multiple theories or perspectives to interpret a single set of data
2. Methodological triangulation: The use of multiple methods to study a single topic
3. Observer triangulation: The use of more than one observer in a single study to achieve intersubjective
agreement
4. Data triangulation: The use of more than one data source (interviews, archival materials, observational
data, etc.)

Valerie Janesick (2000) added a fifth type of triangulation: interdisciplinary triangulation, or using more than
one discipline in a single study. Among these types, triangulation of data is most commonly used as a strategy
for rigor.

From its earliest use solely as a means of confirmation, triangulation has expanded in definition to include
completeness and the enlargement of perspectives (Flick, 2004). Triangulation by theory and triangulation by
discipline are a good fit with this new definition. In contrast, triangulation by observation implies validation or
corroboration. An example of this can be found in Snow and Anderson’s (1991) study of the urban homeless
in which they used multiple observers to enhance the reliability of their observations. Triangulation by
method most often refers to using qualitative and quantitative approaches in tandem. Known more commonly
as “mixed methods,” this type of triangulation has many iterations (more on this in Chapter 10).

Having multiple coders could be considered a form of analytic triangulation. As discussed in Chapter 6,
independent coding and comparison are valuable safeguards against bias in data analysis. Although sometimes
treated as calculable (as in coder agreement rates and Cohen’s kappas), independent coding is better
understood as triangulation for completeness rather than for accuracy.

Triangulation by data source is probably the most commonly used strategy for rigor. When data from field
notes, interviews, and/or archival materials are convergent, one has greater confidence that the observations
are trustworthy. As shown in Figure 8.1, triangulation helps to counter all of the threats to trustworthiness.
Yet, inconsistencies and contradictions are common (Bloor, 2001). When this happens, one must decide
whether to favor one source over another or view discrepancies as an opportunity for new insights. Riessman’s
study of divorcing partners (1990) found that men and women gave very different accounts of their marital
problems. Their conflicting stories became a jumping-off point for a discussion of the ways that men and
women make sense of divorce. Just as negative case analysis may open the door to refining one’s
interpretations, disagreement among data sources may lead to new and expanded perspectives.

With its underlying assumption of a single reality amenable to multiple captures via different data,
triangulation can appear deeply problematic for a constructivist researcher. If one accepts multiple realities and

301
contexts of knowing, then a more simpatico term might be crystallization (Richardson, 2000). In practice,
crystallization brings to bear the same pursuit of diversity in viewpoints, data, and theories as does
triangulation, but the purpose is to amplify not corroborate.

302
Peer Debriefing and Support
Debriefing and support often go unheralded as a means of keeping the “instrument” sharp and true. When
done correctly, it is a type of group reflexivity that gives the researcher fresh perspectives and guards against
bias. Such support may come from an academic advisor or a mentor, but one of the most effective means is
peer debriefing and support (PDS; Padgett, Mathew, & Conte, 2004). Given the growing tendency to
conduct qualitative research in teams, PDS can be built into the study’s infrastructure—it just needs activating
and nurturing. In both the New York Services Study (NYSS) and New York Recovery Study (NYRS), we met
as a team every week, and the agenda invariably included reviewing data collection for the week and any
problems or questions that arose.

Figure 8.1 Strategies for Enhancing Rigor and Trustworthiness

303
PDS groups can be a lifeline for novice qualitative researchers (Steinmetz, 1991). Members get and give
feedback, offer new ideas, and simply recharge their batteries. Attending a PDS group meeting gives the
researcher a chance to share the inevitable ups and downs of fieldwork and data analysis. However, the role of
these groups in qualitative research is not simply socioemotional. They are also a mechanism for keeping the
researcher honest (Lincoln & Guba, 1985). To this end, PDS groups contribute to the rigor of a qualitative

304
study by reducing researcher bias (see Figure 8.1).

PDS group members may share insights or hunches for feedback. They may read portions of their coding
memos along with relevant (and de-identified) chunks of data to see if the codes make sense. Group members
may read passages from their field notes and journals to get reactions from others about their observations and
their ability to be reflexive. Some group members may become reciprocal co-coders. There are rarely any set
rules as long as confidentiality is maintained, and the focus stays on constructive criticism.

PDS groups work best when they meet on a regular schedule and rotate leadership roles among group
members. Their composition can be homogeneous by discipline or they may be multidisciplinary.
Homogeneity allows members to draw on common interests, communicate in a common language, and
reduce time spent negotiating disciplinary boundaries. Heterogeneous groups drawn from diverse fields can be
intellectually invigorating. For qualitative researchers who do not have access to a PDS group, acquiring even
one peer can be helpful. Given busy schedules and travel time, PDS groups can meet online through video or
audio conferencing.

In addition to their rigor-enhancing qualities, PDS groups perform instrumental functions. Group members
give helpful hints, for example, remembering to take extra batteries to interviews. They share news of the
latest in QDA software upgrades and tips on how to use them. Members encourage each other to set and
meet deadlines. They might offer suggestions on how to negotiate with an unhappy partner feeling neglected
by the researcher’s time spent in the field. Having like-minded others helps reduce stress during the rough
passages and celebrate the good times.

PDS groups have a few risks. They can collectively veer off course, either by fostering a sort of “group think”
atmosphere of enforced conformity or by becoming hypercritical and intolerant. New ideas are fragile
creations and can get crushed even by well-meaning colleagues (Wolcott, 2001). When done properly,
however, peer support is an invaluable addition to the repertoire of rigor enhancement.

305
Member Checking
As data collection segues into analysis and interpretation, qualitative researchers may seek verification of
preliminary findings by going back to the study participants. Member checking (Lincoln & Guba, 1985) can be
an important step in guarding against researcher bias. It also represents a logical extension of the close
relationship between the researcher and the respondent. For qualitative studies in which researchers and
participants are co-researchers, member checking is integral to the study’s design.

However, member checking raises a number of questions, making it one of the more problematic of rigor
strategies in theory and practice (Tracy substitutes “member reflections” as a preferred moniker; 2010). Two
problems emerge from this. First is the basic issue of what exactly members are checking. If asked to review
their own interview transcripts or personal case summaries, participants’ authority is rarely challenged. But
member checks often refer to cross-case interpretation; it is reasonable to wonder if an individual respondent
can (or should) be expected to pass judgment on findings involving many hours of immersion and synthesis of
data from multiple participants. Second, member checking (as with triangulation) implies that a single reality
can be captured (albeit this time by participants rather than by peers or others; Sandelowski, 1993). Thus,
participants may disagree with a study’s findings not because the findings are inaccurate but because they have
their own standpoints and realities, all of which may change over time (Rolfe, 2006).

The practicalities of member checking can be daunting. Some participants do not want to be bothered or will
hastily rubber stamp whatever is in front of them. Others may not wish to revisit their earlier statements and
emotions. For some studies, logistical barriers or gatekeeper issues prevent going back—frail elders,
recalcitrant adolescents, and busy executives come to mind. Member checks also fall into a gray area when it
comes to human subjects committee oversight. Can they be considered part of the approved interview
protocols and thus not require additional consent? What happens if and when member checking verges into
new data collection? This is a fine line that often cannot be predicted or managed in advance.

Member checking can prompt revisiting the data and generating new interpretations, but the researcher may
wish to stick to her viewpoint (and explain why she feels this way). While member checking is not feasible for
some studies (Morse, 2015), its underlying premise and intention—to engage participants to react, reflect, and
expand on the nascent findings—is commensurate with the values of qualitative inquiry.

306
Negative Case Analysis
Although discussed in Chapter 7, the importance of negative case analysis bears repeating here (see Figure
8.1). Just as the PDS group challenges a researcher to explore personal biases, negative case analysis puts the
onus on the researcher to operate in a critically self-reflective way. As such, it enhances trustworthiness by
seeking alternative explanations and avoiding lopsided interpretations (Morrow, 2005). Negative cases may
disprove preliminary hunches, or they may turn out to be an anomaly that affirms that the researcher is on the
right track (i.e., the exception that proves the rule). This can only be ascertained by close and frequent passes
through the data. Of course, this presumes entering into deductive thinking during analyses—a common
practice in grounded theory but not necessarily in other qualitative approaches.

307
Auditing—Leaving a Decision Trail
Leaving an audit trail means adopting a spirit of openness and documenting the steps taken in data collection
and analysis (Lincoln & Guba, 1985). The components of an audit trail include samples of raw (de-identified)
data as well as memos and iterative versions of a codebook or thematic analysis plan. Although not intended
to squelch creativity or promote lockstep conformity, audit trailing implies a degree of accountability that
discomfits some qualitative researchers for two reasons, one logistical and the other epistemological. First,
keeping an audit trail adds considerable burden to a study. Keeping track of one’s analytic decisions and
creating an annotated archive of study documents is a layer of effort that can feel like a thankless task,
especially when there are no auditors waiting in the wings. Second, documentation and external accountability
are viewed by some researchers as unwelcome intrusions and indicative of a culture of bureaucratic oversight.

Leaving an audit trail is more often honored in the breach than in reality. In my experience, it is rarely used
except through a voluntary ad hoc process of sharing memos and analytic decisions with one’s advisor, mentor,
or PDS group. Auditing is not practical as a means of evaluating qualitative research by journal peer reviewers
because of time limitations as well as the potential loss of privacy. However, when serious (or even somewhat
concerning) questions are raised about the integrity of the work, requesting copies of memos or de-identified
data seems reasonable. Regardless of whether outsiders have access, all qualitative researchers can benefit from
keeping track of their analytic decisions and organizing the data and memos in a systematic fashion.

308
Strategies for Rigor Applied to the Six Qualitative Approaches
Some rigor strategies are a better fit than others, depending on the type of qualitative approach being used.
Ethnographic research is defined by its reliance on prolonged engagement, but this does not preclude using
the other strategies (member checking and triangulation of data are readily applicable). Case study, grounded
theory, and action-oriented research can benefit from any of the six strategies, with data triangulation leading
the way in case studies and negative case analysis historically allied with grounded theory. Reliance upon texts
within narrative approaches makes data triangulation unlikely and negative case analysis inappropriate. In
phenomenological approaches, peer debriefing and auditing could be considered contaminating influences
that interfere with the search for deep structures of meaning. Member checking and prolonged engagement,
on the other hand, are made easier by the closeness of the researcher-participant relationship.

A rare example of a study that used all six strategies (and documented their use in the write-up) can be found
in Morrow and Smith’s (1995) grounded theory research with sex abuse survivors described in Box 2.2 of
Chapter 2. Prolonged engagement consisted of multiple interviews and over 16 months of interactions with
participants. PDS was built into weekly meetings of the research team. Data triangulation was carried out via
in-depth interviews, videotapes of a 10-week focus group, and documents such as participants’ journals, field
notes, and a self-reflective journal kept by Morrow. Member checking was carried out in the study early on by
including study participants as members of the research team. Auditing was facilitated by detailed analytic and
self-reflective memos. The audit trail consisted of chronological narratives of study activities as well as a list of
the 166 codes that formed the basis of the analyses. Finally, negative case analysis was central to the study.
Acknowledging a “human cognitive bias toward confirmation” (Morrow & Smith, 1995, p. 26), the authors
actively searched for disconfirming evidence, including consulting with participants on discrepant findings.

Most of the six strategies are not a good fit with constructivist qualitative research. Leaving an audit trail
suggests inappropriate accountability, and negative case analysis implies deductive rather than inductive
thinking. Similarly, a definition of triangulation as corroboration presumes a reality that can be captured
rather than multiple realities. Member checking and peer debriefing are relevant but not as checks on bias so
much as collaborative endeavors intended to expand and understand. This leaves prolonged engagement as
most consistent with the constructivist paradigm.

Denzin (2008) has argued that criteria in qualitative research (and presumably strategies for enhancing
credibility and trustworthiness) must be paradigm-specific. As we will see later in this chapter, constructivist
notions of quality have been developed, and there have been concerted efforts to be inclusive of multiple
paradigms in addressing issues of quality.

309
Dilemmas of Rigor and Action-Oriented Research
Community-based participatory research (CBPR) and action-oriented research have an ethos of sharing
control with community members (Israel, Eng, Schulz, & Parker, 2005; Minkler & Wallerstein, 2003) that
can introduce a potential loss of methodological rigor (Allison & Rootman, 1996). Community members
rarely have expertise in research methods or the inclination to accept all of the demands incurred by research
protocols. This tension between rigor and relevance, found in all research, is front and center in CBPR.
Although researchers’ career advancement depends on rigor as much as relevance, community members
understandably may not share this point of view. Dichotomous thinking should be avoided here. Too often
viewed as a zero-sum game, CBPR is more appropriately viewed as a balancing act than an either-or
proposition. All six of the strategies for rigor can be applied in qualitative CBPR, and some, for example,
member checks, are closely aligned with its values.

Although ultimate resolution of the rigor versus relevance tension depends on the particular circumstances of
a study, a few recommendations can be made. First, rigor in CBPR and other forms of action-oriented
research takes dedication, time, and patience. Community members have a right to know and question the
study’s methods, if not actively participate in their development and implementation. If and when community
members seek to veto some aspect of the proposed study, the researcher must find a way to either dissuade
them respectfully or accede to their wishes and try another route. Second, the most demanding and distancing
research designs are a difficult fit with CBPR, especially in the early stages when community buy-in is critical.
Thus, an experimental trial to test an innovative program is far more likely to be successful if implemented
after an extensive period of community involvement and collaboration. Finally, threats to the rigor of a CBPR
study are greatest when time is short and results are needed quickly. Given the need for flexibility and options,
researchers in a hurry should probably avoid the CBPR route.

310
Appraising Quality: The Rise of Checklists, Criteria, and Guidelines
Over the years, a number of checklists and guidelines have emerged to guide peer reviewers charged with
appraising the quality of qualitative studies. An outgrowth of the explosive increase in the number of
qualitative studies submitted for publication and funding, the move toward uniform standards gained impetus
as meta-syntheses and systematic literature reviews generated conclusions from qualitative as well as
quantitative studies (Dixon-Woods, Booth, & Sutton, 2007; Popay, Williams, & Rogers, 1998). Meta-
syntheses originated in the expansive landscape of health-related research then extended to studies in social
welfare, criminal justice, and behavioral therapy (Littell, Corcoran, & Pillai, 2008). Following the lead of
quantitative meta-analyses, studies are rated methodologically before the aggregation begins, and weaker
studies are given less weight (or excluded).

Methodological criteria specific to qualitative research can be found in health care (Cohen & Crabtree, 2008;
Inui & Frankel, 1991; Mays & Pope, 2000; Tong, Sainsbury, & Craig, 2007), nursing (Sandelowski &
Barroso, 2002), and social work (Drisko, 1997). The social sciences—sociology and anthropology—have not
focused on producing standardized criteria, and many qualitative methodologists have been cautious (Altheide
& Johnson, 1994; Seale & Silverman, 1997) if not dismissive of “criteriology” (Schwandt, 1994).

However, this has not slowed down the pursuit, especially among qualitative researchers in health-related
professions. Tong et al. (2007) conducted a meta-synthesis of the literature on qualitative method checklists
and retrieved 22 examples containing 76 discrete items. Grouping the items and eliminating redundancy, the
authors came up with a 32-item Consolidated Criteria for Reporting Qualitative Studies (COREQ) measure
consisting of three domains (research team and reflexivity, study design, analysis and findings). The domains
focused on personal characteristics of the researchers and their relationship with participants (Domain #1);
theoretical framework, participant selection, research setting, and data collection (Domain #2); and data
analysis and reporting (Domain #3).

Cohen and Crabtree (2008) conducted a similar review of the literature and identified 29 journal articles and
16 books or book chapters that offered explicit criteria for evaluating the quality of qualitative research. Seven
evaluative criteria were distilled from these works: (1) carrying out ethical research; (2) noting the importance
of the research; (3) clarity and coherence of the report; (4) use of rigorous and appropriate methods; (5) use of
researcher reflexivity; (6) establishing validity or credibility; and (7) establishing verification or reliability. As
with the Tong et al. (2007) effort, these criteria are an epistemological hybrid. The authors noted that there
was less agreement on the last three of the criteria (Cohen & Crabtree, 2008).

Admittedly, any attempt at synthesizing is a challenge. It is not easy to maintain an allegiance to quality
standards without sacrificing the curiosity, creativity, and serendipity that make qualitative research distinct
and distinguished. Indeed, voices from within the evidence-based movement have made this point. Barbour
and Barbour (2003), citing the seminal work of Greenhaugh (1997), argue for a less reductionist approach to
synthesis. They go on to note that many if not most qualitative studies previously published would have to be
eliminated under such rising expectations. Eakin and Mykhalovskiy (2003) sum up these concerns, “gauging

311
quality primarily in terms of the execution of proper method, guidelines tend to oversimplify and
inappropriately standardize the complex and non-formulaic nature of qualitative inquiry, promoting the
notion of a fixed relationship between research practice and knowledge generated” (p. 192). Thus, a fixation
on “procedural correctness restricts the reader’s field of vision on the research process and diverts attention
away from the analytic content of the research” (p. 192).

Less concerned with the pursuit of criteria to enhance evidence-based evaluations, some constructivist and
postmodern researchers emphasize esthetic and emancipatory aspects of qualitative research rendered as
performance, poetry, fiction, and drama. Denzin and Giardina (2006) recommend, “a methodology of the
heart, a prophetic, feminist post-pragmatism that embraces an ethics of truth grounded in love, care, hope and
togetherness” (p. xvii). Opposition to the elitism and conservatism of scientific methods animates such
poststructural romanticism (Hammersley, 2008). Yet, as mentioned earlier in this chapter, the substitution of
esthetics for criteria-based standards opens the door to elite appraisals as carried out in the art world, not the
worlds of the social sciences and the professions. If quality is only in the eye of the beholder, more powerful
beholders will hold sway.

Charmaz (2014) adopts a middle ground, acknowledging the need for esthetic criteria but distinguishing them
from methodological concerns. In doing so, she offers 19 criteria consonant with constructivist grounded
theory grouped under the headings of credibility, originality, resonance, and usefulness. Credibility criteria
include (1) intimate familiarity with the setting and topic; (2) sufficient data; (3) systematic comparisons; (4)
wide-ranging categories; (5) logical links between data, analyses, and findings; and, (6) claims are grounded in
sufficient evidence for the reader to agree with them. Originality is evidenced in (7) fresh categories and
insights; (8) new conceptual renderings; (9) social and theoretical significance; and (10) challenging,
extending, or refining existing ideas, concepts, and practices. Resonance refers to (11) categories that capture
the fullness of the studied experience; (12) attention to liminal and unstable taken-for-granted meanings; (13)
links are made between larger institutions and individual lives (when needed); and (14) findings make sense to
participants. Finally, usefulness criteria are met when (15) interpretations are usable in everyday worlds; (16)
analytic categories suggest generic processes; (17) generic processes are examined for tacit implications; (18)
analyses spark further research; and (19) the work contributes both to knowledge and a better world. Though
expansive and open to differing readings, Charmaz’s criteria capture basic principles of credibility, creativity,
and utility.

312
In Search of the Big Tent: Balancing Rigor and Relevance
Invoking the importance of relevance requires a sense of what it means and how it is derived. In the abstract,
relevance refers to making connections. In qualitative research, audiences relate to the findings when they find
value in the knowledge produced. Social work and other practice-based researchers are attuned to issues of
inequality, social exclusion, and disempowerment, and this is where relevance needs to have traction. The
comfortable remove of academia has to be left behind to engage in important issues of the day. To limit one’s
pursuit of relevance to personal experience or to a few like-minded others is to risk foregoing impact in the
worlds of policy and practice.

Along with the study’s methodological rigor, the relevance of a qualitative study is assessed by peer reviewers.
While admittedly emphasizing rigorous methods, virtually all of the quality checklists make a point of
evaluating the study’s relevance. After all, “part of making scholarship powerful is talking in ways that are
appreciated by a variety of audiences, including grant agencies, governmental officials, and media contacts—
many of whom are unfamiliar with the methodology of qualitative research” (Tracy, 2010, p. 838).

In a widely cited attempt to bridge epistemological and methodological differences, Sarah Tracy (2010)
identified eight “big tent” criteria for quality in qualitative research. Rejecting Denzin’s (2008) argument that
standards should be paradigm specific, she notes that beneath the various paradigms is a substrate of craft
skills commonly used by qualitative researchers. The eight markers of goodness are (1) worthy topic, (2) rich
rigor, (3) sincerity, (4) credibility, (5) resonance, (6) significant contribution, (7) ethics, and (8) meaningful
coherence. While having a worthy topic hardly requires explanation, it merits inclusion as that nagging “so
what” question awaits an answer. Rich rigor—a seeming oxymoron to some—covers the use of theory as well
as collection of sufficient data in the right contexts. Sincerity combines reflexivity and transparency, thereby
binding together two central concerns in qualitative methods. Credibility rests on the use of triangulation (or
crystallization), thick description, multivocality, and member reflection (similar to member checking).
Resonance refers to findings that are evocative as well as transferable. Significant contribution means that the
study has made its mark either in new theoretical ideas, methodologically, or practically. Ethics covers all
aspects of protecting one’s participants including a priori approvals, in-progress monitoring, and exiting the
study respectfully. Finally, meaningful coherence is an ideal state of affairs signified by the study’s ability to
accomplish what it purported to. Success in this regard refers to meaningfully interconnecting literature,
research questions, findings, and interpretations.

To be sure, these criteria are largely dependent on subjective judgment. How do we distinguish thick from
thin description? Crystallization from triangulation? Coherence from incoherence? While most can be applied
directly to the qualitative report, others might need more time to ripen and be made known, for example,
significant contribution and transferability. Upon closer examination, one can find several of the strategies for
rigor within the eight criteria—triangulation, member checking, prolonged engagement—and peer debriefing
is consonant with credibility as described. Similarly, an audit trail might be one means of achieving
transparency and sincerity. Negative case analysis is the least visible, but it is not uniformly practiced or
expected. Other criteria mentioned earlier in this chapter can be found resting comfortably inside the big tent,

313
for example, ecological validity and evidentiary adequacy.

Taken together, “big tent” criteria are intended to address both rigor and relevance, at least in the short term.
How does one pay heed to these or any criteria? First, qualitative researchers can include them as an integral
part of the study’s design and planning. Second, one can refer to criteria while documenting what actually
transpired in the study to assure readers that guidelines were followed. Qualitative studies entail an obligation
to report what happened after the fact since flexibility leads to alterations in the conduct of the study.

Qualitative researchers have grown fond of publishing behind-the-scenes accounts of their studies (see
Padgett, 2004a for an edited compendium of social work qualitative studies). Such openness serves two
worthy purposes: reflexivity and transparency. Providing what amounts to a readable audit trail engages
readers by taking them backstage in sharing the ups and downs that invariably took place. It is reflexivity in
action (in writing).

Behind-the-scenes portrayals also help explicate the methods and make them more accessible. Social work
researcher Glen Bowen (2006) has been especially adept at this, drawing on his grounded theory study set in a
community-based antipoverty project in Jamaica. Following a constructivist paradigm, he applied sensitizing
concepts such as empowerment and social capital while developing a grounded theory of stakeholder
collaboration in development work. Bowen has also written about his use of saturation (2008), document
analysis (2009), and ways to publish from one’s dissertation (2010). This use of self, central to reflexivity, has
become a means of sharing experiences but also providing guidance for others (Gair, 2002). A take-what-is-
useful-to-you attitude is characteristic of qualitative inquiry and its nonregulatory bent.

314
Summary and Concluding Thoughts
By now, the reader may be wondering if any stone has been left unturned in this chapter’s discussion of what
makes a trustworthy or high-quality study. All can be summarized in a few interrelated points: (1) quality is
important in qualitative research, yet consensus on standards remains elusive; (2) evidence of quality emerging
from a completed study is not the same as building into a study’s design deliberate efforts to improve quality;
(3) achieving rigor can be enhanced by using one or more of six strategies developed explicitly for qualitative
studies; (4) guidelines and checklists for appraising quality have gained momentum in recent years; and (5)
consternation over strategies for rigor and uniform standards has given rise to a “big tent” approach as most
likely to garner support from qualitative researchers across the paradigmatic spectrum.

The capacity for a study to stimulate thought, improve practices and policies, and incite further research is a
metric of success agreeable to most anyone. And, for good or ill (and sometimes both), peer reviews serve a
gatekeeping function in determining the quality of research. Having no criteria or depending upon a default
form of art world elitism can lead to confusion and inhibit the dissemination of qualitatively derived
knowledge. Despite the complicating factors and demands, rigor in some form and relevance in multiple forms
are desirable goals for qualitative researchers.

315
Exercises
1. This can be done alone or in a group. Select and read a qualitative study from one of the following
journals: Qualitative Health Research, Qualitative Social Work, or Qualitative Sociology. Note what (if any)
strategies for rigor the authors mention in discussing their methods and findings.
2. Read the article titled “Reading Qualitative Studies” by Sandelowski and Barroso (available online in the
International Journal of Qualitative Methods, 1(1), 2002,
http://www.ualberta.ca/~iiqm/backissues/1_1Final/html/sandeleng.html). Following their guidelines,
choose a published study or a dissertation, and critique it. What are some of the topics most often
neglected by authors?
3. In a group, discuss the three threats to trustworthiness (reactivity, researcher bias, and respondent bias)
and how they differ in quantitative versus qualitative studies. Are quantitative studies more vulnerable to
some threats than others? What about qualitative studies?
4. Debate the pros and cons of having checklists for appraising qualitative studies. Among the criteria
mentioned in this chapter, which ones appear most applicable to all qualitative studies?

316
Additional Readings
Altheide, D. L., & Johnson, J. M. (1994). Criteria for assessing interpretive validity in qualitative research. In
N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 485–499). Thousand Oaks, CA:
Sage.

Davies, D., & Dodd, J. (2002). Qualitative research and the question of rigor. Qualitative Health Research,
12(2), 279–289.

Drisko, J. (1997). Strengthening qualitative studies and reports: Standards to enhance academic integrity.
Journal of Social Work Education, 33, 187–197.

Kidd, P. S., & Parshall, M. B. (2000). Getting the focus and the group: Enhancing analytical rigor in focus
group research. Qualitative Health Research, 10(3), 293–308.

LeCompte, M. D., & Goetz, J. P. (1984). Problems of reliability and validity in ethnographic research.
Review of Educational Research, 52, 31–60.

Lincoln, Y. S. (1995). Emerging criteria for quality in qualitative and interpretative research. Qualitative
Inquiry, 1(3), 275–289.

Marshall, C. (1990). Goodness criteria: Are they objective or judgment calls? In E. G. Guba (Ed.), The
paradigm dialog (pp. 188–197). Newbury Park, CA: Sage.

Maxwell, J. A. (1992). Understanding and validity in qualitative research. Harvard Educational Review, 62(3),
279–300.

Mays, N., & Pope, C. (2000). Qualitative research in health care: Assessing quality in qualitative research.
British Medical Journal, 320, 50–52.

Morrow, S. L. (2005). Quality and trustworthiness in qualitative research in counseling psychology. Journal of
Counseling Psychology, 52, 250–260.

Morse, J. M., Barrett, M., Mayan, M., Olsen, K., & Spiers, S. (2002). Verification strategies for establishing
reliability and validity in qualitative research. International Journal of Qualitative Methods, 1, Article 2.

Sandelowski, M. (1993). Rigor or rigor mortis: The problem of rigor in qualitative research revisited. Advances
in Nursing Science, 6, 1–8.

Seale, C. (2002). Quality issues in qualitative inquiry. Qualitative Social Work, 1(1), 97–110.

Tracy, S. J. (2010). Qualitative quality: Eight “big tent” criteria for excellent qualitative research. Qualitative
Inquiry, 16, 837–851.

Whittemore, R., Chase, S. K., & Mandle, C. (2001). Validity in qualitative research. Qualitative Health

317
Research, 11(4), 522–527.

318
9 Telling the Story Writing the Qualitative Study

Qualitative researchers who have a talent for writing are fortunate souls. For most, however, writing up a
qualitative study is a craft to be learned and honed over time. Susser (1997) drew a distinction between
“authoring” and “writing.” Thus, many journal articles and books are authored (and coauthored), but not all
authors are writers. Writing is not merely reporting, it is art and craft. To illustrate this point, consider how
most quantitative reports look. Full of tables and graphs, their readability takes second place to numerical
precision. The terminology of statistical analyses, embedded within a standardized reporting format, produces
a writing style that informs but rarely captivates.

If the act of writing up quantitative research is pro forma and somewhat anticlimactic (a well-conducted study
can usually survive a poor write-up), it is the climactic event in qualitative research (Padgett, 2004b).
Disseminated studies are “vast conversations with dispersed others” (Flaherty, 2002, p. 510), and qualitative
researchers have an obligation to ensure the conversation is lively and thought provoking. The difference
between a successful report and a ho-hum report lies in how well the salience of the findings shines through
(Becker, 1998).

The ideal scenario in writing up a qualitative study involves maintaining alignment between one’s choice of
method and the terminology used to describe the study and its procedures. Sandelowski and Barroso (2003)
examined a number of qualitative studies and found this ideal was rarely attained. Instead, authors obscured
the diversity of their approaches through generic description, many relying on what appeared to be variations
of content analysis. For studies that offered specifics, there was frequently a disconnect between their self-
labeling and reality, for example, a study that was ostensibly phenomenological used a watered-down version
of grounded theory.

Two rather obvious suggestions can help prepare for writing. First, read as many good qualitative studies as
possible. Exemplary articles have been cited throughout this book as well as full-length monographs. Reading
and absorbing both the form and substance of qualitative studies will expose you to the many options available
for the study write-up and help you to develop your own style. Second, begin writing early, and build in plenty
of time for editing and polishing. To employ a sports analogy, if reading is like being a spectator, writing is
like being a player (Lofland & Lofland, 1995). It is better to write even when you do not feel like it than to
wait for a lightning bolt of inspiration to strike (Wolcott, 2001). Start with an outline, and flesh it out as you
go. Difficulties in writing are less often due to “writer’s block” than to “idea block” (Lofland & Lofland, 1995,
p. 205). The self-selection process leading individuals toward qualitative research tends to favor creative,
abstract thinkers. The challenge is to keep those ideas and abstractions grounded in the data.

319
Preparing to Write: Key Decision Points
Many qualitative researchers have shared their own strategies for writing, including Gilgun (2005), Holliday
(2007), Richardson (2000), van Manen (2006), Weiss (1994), and Wolcott (2001). In reviewing their works
and through my own writing experience, I have identified two key decisions all qualitative researchers should
consider when approaching the write-up phase. These are the audience for the study and the researcher’s role
in the report. Two additional decisions are pivotal for some but not all qualitative researchers: whether and
how to use numbers and authorship.

320
Identifying the Audience
Qualitative studies may be written up for a number of purposes and audiences. For academic audiences—
doctoral dissertation committees, journal reviewers, and conference attendees—the write-up is pitched at a
high level to ensure that the knowledge being reported has conceptual and theoretical heft. Evaluative reports
are written in more practical language and are structured to provide concrete suggestions for improving
policies and practices.

When a study is commissioned and funded, the sponsoring organization usually wants the researcher to assess
problems and recommend solutions. The widest possible audience—the general public—is one that some
academic researchers scorn and others long for. The benefits of reaching a popular nonacademic audience lie
in providing an accessible entrée to one’s research topic. Qualitative studies have little distance to travel in this
regard. Several of the monographs listed at the end of Chapter 1 have enjoyed wide distribution and far-
reaching influence.

Audiences can be both proximal and distal. Closest in proximity are one’s study participants, and their access
to study findings has become integral to the values of qualitative inquiry. More distant audiences of
policymakers and practitioners are aspired to when one’s work is designed to have an impact on people’s lives
(social work researchers fall squarely into this group). Distal audiences are reached online—researchers,
students, and other information-seekers turn to Internet search engines armed with a few key words and
intellectual curiosity. Though dispersed in space and time, this audience can be the ultimate arbiter of success
through downloads of journal articles as well as social media call-outs.

Metrics are ever present, for example, the impact factor score of a journal, the number of citations, first versus
second or lower rank in authorship. The value placed on quantifying impact, as any qualitative researcher
knows, can produce more of a popularity contest than an evaluation of intellectual heft. Moreover, impact may
need to gather steam over time as audiences find new ways to relate to a study’s relevance. A final point is that
desired audiences may be multiple, and this may require writing different reports for the same study. The
more that a qualitative report reverberates in the world of ideas as well as in praxis, the more likely it will have
many audiences.

321
The Researcher’s Role in the Report
The de facto stance for a qualitative researcher up to the 1970s was that of all-knowing yet invisible outsider,
what Haraway (1988) refers to as the “God” trick.” Termed “Realist Tales” by Van Maanen (1988), such
works took scant notice of the researcher—to comment on that role was an unthinkable breach of etiquette.
Authors of Realist Tales spent many private hours regaling colleagues and students with stories of their
experiences in the field but then edited such stories out of the final report as unseemly self-indulgence (Van
Maanen, 1988).

The rise of constructivism and postmodernism directly challenged this state of affairs. To postmodern critics,
the human emotions stirred by fieldwork were squeezed out of the report by a “Doctrine of Immaculate
Perception” (Van Maanen, 1988, p. 73). In the meantime, academic and general audiences have come to
accept the use of first-person pronouns so it is easier to bring in the researcher’s role. In some quarters, the
pendulum has swung all the way to the researcher as the focus of interest with auto-ethnographies, memoirs,
and other “Confessional Tales” (Van Maanen, 1988).

The researcher should not be edited out of the final report, but that begs the question of how prominently or
subtly this role should be featured. Much of one’s experience in the field—failures as well as successes—can be
pertinent to the credibility of the study. The decision about how much personal information and experience to
include in the report is influenced not only by one’s epistemological stance but also by the intended audience
and the outlet for its dissemination. Where some audiences appreciate full reflexivity, others might find it self-
indulgent or distracting.

The form reflexivity takes can range from a brief mention to a thorough description of how the researcher
entered the field, interacted with participants, and grappled with potential intrusions of personal bias. The
latter, often drawn from the personal journal or diary kept during the study, situates the author as an active
participant in the study. Journal articles tend to constrain being self-reflexive in part due to the shortage of
space and the need to give the methods and findings their full due. Long-form reports (e.g., dissertations and
books) allow more space for the researcher’s role threaded into the narrative or described in detail in an
appendix.

322
Using Numbers in the Report
The decision about whether and how to use numbers in the final report is not cut and dried (Maxwell, 2010).
To be sure, qualitative researchers do not give much credence to numbers and the use of some, many or few is
the norm (Weiss, 1994). On the other hand, a table showing participant demographic characteristics depends
upon numbers or proportions, and the results section of the report is where numbers can sometimes be helpful
if used judiciously and without disrupting the narrative. In a study of caregiving daughters and their elderly
mothers, Walker and Allen (1991) identified three relationship types (intrinsic, ambivalent, and conflicted)
and the percent of mother–daughter pairs for each type (45%, 34%, and 21%, respectively). Providing these
proportional indices helps the reader to understand how the types were distributed among the mother–
daughter pairs who were studied.

The problematic aspects of using numbers should not be ignored. First, numerical findings can give a false
impression of precision where none exists. Thus, “most of the women in the study expressed satisfaction with
their medical care” sounds better than “80% were satisfied with their medical care” when “80%” refers to 8 out
of 10 women and “satisfaction with medical care” was not measured or calibrated. At the same time, numbers
are sometimes useful when presenting findings that are clear-cut, for example, “10 agency directors (40% of
those who participated in the study) lost their jobs during the course of the study.” Another caveat is that
many important findings in qualitative studies were not routinely elicited but emerged serendipitously and
voluntarily. As such, any count of the number of respondents who gave certain information is probably an
undercount (a numerator without a denominator) and is thus misleading for those who assume it is a true
measure of frequency.

Here is one use of numbers the researcher might consider in the methods section: reporting the number of
pages of transcripts, number of hours spent in the field, and the number of hours spent conducting interviews.
In her doctoral dissertation conducted as part of the New York Recovery Study (NYRS), Emmy Tiderington
(2015) noted the following in her methods section: “In total, 106 hours were spent in the field resulting in 164
pages of field notes collected by five different NYRS team members, including myself” (p. 61). The number of
pages of transcripts alerts the reader to the time required to read and code—this would give any quantitative
researcher pause. For example, a qualitative study with a sample size of 20, two interviews per participant, and
at 60 to 90 minutes per interview would conservatively yield 1,000 pages (40 interviews at an average of 25
pages per interview). Of course, too much of this type of information can be overkill, but not enough
information is a worse offense.

323
Authorship and Coauthorship Decisions
Qualitative research has a long and honorable tradition of solo work. A trend toward teamwork, however, has
made coauthorship common and expected (even doctoral dissertations may be transformed into coauthored
manuscripts later on). This trend has been especially evident in professions where research collaboration was a
traditional byproduct of working in multidisciplinary settings. The egalitarian ethos of qualitative inquiry
affects the teamwork, making leadership roles less foreordained and subverting attempts to impose decisions
based on seniority.

Qualitative collaborations present challenges that quantitative researchers do not normally encounter.
Qualitative studies are multifaceted and interwoven—every phase informing and being informed by every
other. Viewed as a series of feedback loops, qualitative designs unfold in ways that are not always predictable.
Team members’ skill sets need to overlap rather than simply complement one another. This seamless quality
presents challenges when writing the report.

Deciding on authorship can be a head scratcher. First, there is the decision about who is and who is not an
author. These days, peer-reviewed journals expect active involvement (no free rides), and simply helping in
some way without being instrumental during the study calls for acknowledgement, not authorship. Beyond
this minimal threshold, there are no hard and fast rules, and much depends upon consensus and a sense of
fairness. A team member may have helped with one phase of the study but not with other phases. Another
team member may have been involved in the study in all phases. Yet another may have started with the study,
moved to another city, but wants to help write the report. These are a few of the possibilities to consider.

The optimal approach happens when a coauthor is involved in key aspects of the study (not necessarily all)
and is involved in the write-up. Divvying up manuscript writing into sections sounds democratic, but it rarely
works out well—with multiple authors much time can be spent negotiating the blending of the parts. In the
NYRS, we started with deciding on who would be the first author, that is, the team member who would own
it and coordinate the writing from that point. The first author can seek out team members to write specific
sections that they are more knowledgeable about especially if a section, for example, sampling procedures, is
fairly compartmentalized and can be easily dropped in. But the first author has the job of making the report
cohere.

The introduction and discussion sections are where collaboration and multiple perspectives can shine through.
In the NYRS, the first author usually took a stab at the literature review and then invited others to give
suggestions for additional citations or ideas left out. If the basic framing of the problem looks weak, it may be
time to go back to the drawing board. Manuscripts invariably go through multiple drafts so patience is a virtue
here. The discussion section is also an arena for coauthor input, but it needs to segue smoothly from the
earlier sections (and circle back to the original research questions).

As the director on the New York Services Study (NYSS) and NYRS, I wanted to be first author on the one or
two main outcome papers (in keeping with the funder’s expectation but also my own sense of entitlement),

324
but after that, other papers were open for first authorship. I would sometimes referee first-authorship access
among team members, but the group bonhomie and an abundance of data made differences in opinion few
and easily resolved. In our studies, there were two different participant groups—consumers (formerly homeless
men and women) and case managers who worked in the housing programs where the consumers were living
and receiving services. In the NYRS, two team members—Emmy Tiderington and Mimi Choy-Brown—
were seasoned social workers and doctoral students with a keen interest in social work practice and
organizations. They, along with a faculty member with expertise in this area (Victoria Stanhope), owned the
case manager data, taking responsibility for its collection and analysis. The project director (Bikki Tran
Smith) and I focused on the consumer data including the photo-elicitation interview component. Our
colleague Ben Henwood was an all-purpose team member interested in both sources of data (until he
graduated to become a professor far away).

Deciding on authorship after the first author can be a sticky wicket. In my experience, there is a big drop in
effort starting with the second author. The opposite situation—where the second author assumes most or all
of the work of the write-up and a senior team member claims first authorship—should be a thing of the past.
This still leaves questions to be answered. Should authorial rank be strictly apportioned according to who did
what and how much? How is this ascertained and decided? What if a team member is a poor writer or
faltering in time management? What if a team member is going up for tenure and asks to be given authorship
priority? What if a colleague offers some ideas over lunch then asks to be included as a coauthor? The medical
model tradition is one of reserving last authorship for the most senior team member—a more coveted position
than being in the middle but lower than first author. Otherwise, coauthorship down the line is apportioned
according to contribution and usually negotiated during or after the work has been done.

Meanwhile, single-authored studies are in no danger of disappearing (some tenure and promotion committees
expect this as well as coauthorship). For researchers seeking recognition for their ideas as well as their
analyses, this can point to a book or to journal articles depending upon disciplinary and institutional
expectations. Peer-reviewed journal articles are more commonly expected for hiring, tenure, and promotion in
academic institutions (outside of some social science departments where books still matter). Coauthorship for
a journal article is the norm, but one’s publishing portfolio should manifest some solo outings.

325
Authorship in and From Dissertations: A Negotiation With Few Rules
If authorship decisions come with few rules and many opinions, coauthorship from dissertations is an
uncharted frontier. Albeit much less common compared to large-scale quantitative studies, qualitative
dissertations may be embedded within a larger research project. In this instance, carving out what is owned by
the doctoral student versus the parent study requires ongoing negotiation. It can be done, although the fact
that the student is in a less powerful position makes it incumbent on the director of the parent study to ensure
the student’s needs are met.

The customary scenario—where there is no parent study—affords more options for the doctoral student
although even here authorship questions can come up. For example, the dissertation advisor or another
committee member may have been drawn into the research in a substantive role such that it merits a
discussion of coauthorship after graduation. A strong dissertation can usually yield up to three journal articles
more or less. Typically, the student single-authors the main findings, then the committee member enters as
coauthor on the other manuscripts. Decisions on apportioning the dissertation for manuscripts should be left
to the student.

The main challenge is deciding how to break up the unified whole of a qualitative dissertation into
publishable chunks. All that effort spent on making it cohere as only a qualitative dissertation can—and now
having to cut it up! Sometimes, one manuscript is devoted to the theoretical or conceptual framework or
literature review, although journals tend to shy away from nonempirical submissions. An especially rich
dissertation can spin off articles on new or unanticipated topics that can constitute a separate article with some
additional analysis, literature review, and write-up.

Then there is the three-article dissertation format that is gaining in popularity. Unlike the monograph format,
the student is required to produce three publishable journal articles. This is clearly the product of a
quantitative mindset, but it has become widely applied to qualitative dissertations as well. The usual
expectation that the student will embed the three articles within a structure—an introductory chapter and a
conclusion chapter tying the work together—makes this an amalgam of sorts.

Producing discrete and identifiable articles makes publishing easier, but authorship questions must come up
earlier. Because some doctoral programs require publishing one or more articles before graduation, the student
must be able to master manuscript writing and formatting as well as the rules of journal submission. For those
new to this enterprise, an advisor or other faculty member may get involved to the extent of earning
coauthorship. This is not unusual, but it again raises concerns about the student’s weaker negotiating position
vis-à-vis a faculty member. Given all of the work that a dissertation takes, erring on the side of the doctoral
student/graduate in deciding authorship seems more than reasonable.

326
Organizing the Report—Key Components
Now that we have addressed some of the preparatory work before writing, the next step is to put pen to paper
(or, more likely, fingers to keyboard). Despite the creative latitude afforded the writer, most qualitative reports
roughly follow the same format as quantitative ones even though their content is very different:

Title. The goal here is to include as many key words as possible while still being catchy (but not
frivolous). It is also important to include qualitative or other descriptor in the title to ensure the report is
retrievable as well as recognizable for its methods. Many authors like the two-phrase option to maximize
description (one phrase descriptive and the other more businesslike). Here is an example: “Innovative
Community Services for Rape Victims: An Application of Multiple Case Study Methodology”
(Campbell & Arens, 1998). Here is another using participants’ own words: “Hell I’m an Addict, but I
Ain’t no Junkie”: An Ethnographic analysis of Aging Heroin Users” (Boeri, 2004).
Abstract. Depending on the publishing outlet, this may be as few as 50 or as many as 250 words. The
abstract may be broken up into the sections described below, or it may be a more holistic narrative.
Either way, one should include the background of and rationale for the study and then summarize the
methods and findings before drawing the conclusions.
Background and Theoretical Context. Here, the author describes the phenomenon of interest including a
rationale for the study and its use of qualitative methods. The literature review that follows places the
study within a theoretical and empirical context. The review may separately cover qualitative studies to
further situate the study and its potential contribution. Theoretical influences and conceptual
frameworks are presented along with the requisite caveat that they are not driving or constricting the
study. The literature review is a sustained argument. This should not lead to unwarranted criticism
(Silverman, 2006). Critiquing the methods and conclusions of empirical work is fair game, but the
argument and rationale is more about gaps in knowledge than errors or incompetence in previous
research. This section usually ends with the research questions.
Methods. The first part introduces the reader to the specific approach (grounded theory, ethnography,
narrative analysis, etc.) with citations of methodologists’ published work in the area. The second part
gives a detailed description of how the study’s goals were accomplished (which often diverges from the
original proposal or plan). Here, the author describes the hows, whens, and whys of the study: its site
and sample selection; entering the field and establishing rapport; data-collection procedures; storing and
managing data; data analyses; human subjects and ethical considerations; and which strategies were used
to enhance rigor. The author should be candid about the limitations of the study and his or her role as
the instrument of data collection and analysis.
Findings. This section goes to the heart of the matter. Qualitative researchers may use a variety of means
to report findings, ranging from graphic displays such as charts, matrices, and maps to lists or schemas
of codes and themes (usually illustrated by selected quotes from the transcripts). They may also present
case vignettes or typologies to illustrate major categories or themes. There are many choices available;
the only requirement is that the findings are understandable and address the original research questions
(with ample room for new or serendipitous findings).

327
 Conclusions and Recommendations. Highlights of the study bear repeating here to remind the reader of
the study’s goals, to summarize how these goals were achieved, and to candidly discuss the study’s
limitations. The implications of the findings are also important. How do they advance knowledge?
What are their applications to practice and policy? Finally, what are suggested directions for future
research?

Here are a few additional pointers: (1) make clear distinctions between content under these four sections, for
example, do not introduce new findings in the conclusion section; (2) use headings and subheadings to
structure the report and to serve as signposts leading the way; (3) cite your sources liberally and often—it is far
better to overcite than to risk even the appearance of plagiarism; and (4) aim for maximum transparency. As
discussed in Chapter 8, transparency and thoroughness in describing one’s methods and findings enhance
confidence in the study and help with the larger and much-needed task of demystifying qualitative research.

328
Qualitative Reports Across the Diverse Approaches
Each of the qualitative approaches has its own writing conventions that represent variations on the format
previously described, but the majority of qualitative reports rely on narrative with the option of including
tables, figures, schemas, photographs, and other illustrative graphics. Grounded theory reports typically follow
the architecture of the theory or conceptual framework that emerged from the analyses. Thus, categories or
themes are listed along with selected quotes to illustrate them. To follow the method to its logical conclusion,
the theoretical schema is diagrammed to visually depict the categories and the relationships between them.

Phenomenological reports tend to follow their own analytic structure. Colaizzi (1978) suggests including
tables of themes and significant statements, a format that Beck (1993) followed in her phenomenological
study of postpartum depression. Moustakas (1994) gives a detailed set of guidelines for phenomenological
reports, including recommending an autobiographical account at the beginning. The Journal of
Phenomenological Psychology contains many examples of this approach.

Ethnographic studies have a tradition of book-length reports, their length deemed necessary to capture the
breadth and depth of the inquiry, but journal articles have become increasingly common. Early ethnographies
had missing or severely truncated sections on methods, but much has changed over the past 20 years as
prominent ethnographers have stepped forward to urge clear explication of methods (Bernard & Gravelee,
2014; LeCompte & Schensul, 1999; Tedlock, 2000)

Narrative approaches often use the zoom in–zoom out technique (Czarniawska, 2004) of spotlighting portions of
annotated texts and then interpreting their meaning according to the study’s specific method (narrative
analysis, conversational analysis, or discourse analysis). Angell and Bolden (2015) featured lengthy annotated
sections of conversations between psychiatrists and patients in their discourse analysis. Published examples of
narrative analyses show how the authors alternate interpretation with relevant interview excerpts (Hyden &
Overlien, 2004; Riessman, 2008; Sands, 2004). By way of contrast, grounded theory and phenomenological
reports place greater emphasis on interpretive structures and illustrative excerpts.

Case studies may vary considerably in format depending on their design (single versus multiple case) and aims.
Stake (2005) describes the basics of a case study report including a description of the study’s purpose
(especially the significance of the case or cases), methods, and setting. Because many case studies are used for
evaluative purposes or to critically analyze an issue, their write-ups typically include typologies, vignettes, and
graphic displays to illustrate the main points and keep the reader from getting lost in the minutiae of the
case(s).

329
Balancing Description and Interpretation
The reader has probably surmised by now that a prime source of variation in qualitative reports is how they
balance description versus interpretation. Although “analytic excess” (Lofland, 2002, p. 158) robs the report of
the rich detail on which it is based, too little interpretation reduces its impact. Concerned about the latter,
anthropologist Clifford Geertz (1973) introduced the term thick description to invigorate ethnographic
interpretation. The phrase has since been widely invoked (and sometimes misunderstood as “more
description” rather than “more interpretation”). Whereas description is localized, interpretation is abstraction,
conceptualization, and broader meaning-making. In interpretation, theoretical lenses are often (re)introduced
to refract the data and findings.

Description still plays an important role in qualitative reports, especially when the topic is new and unfamiliar.
Of course, the decision about what to describe (and how to describe it) can be seen as a tacit form of
interpretation. The anthropologist Margaret Mead (1928), for example, was accused of selectively reporting
what she “saw” while in Samoa in order to fit her preconceived notions about free-wheeling adolescent
sexuality. Ethnographies tend to blend description with interpretation in a seamless manner, but most
qualitative reports present the interpretive structure of the findings interspersed with excerpts of raw data.

The length of the raw excerpt depends on the approach taken by the researcher (care should be taken to avoid
breaching confidentiality when making this decision). Case vignettes tend to run longer to allow full
appreciation of the complexity of the case (Spalding & Phillips, 2007). In the NYRS, we analyzed life history
interviews and reasoned that the only way to illustrate cumulative adversity was through case vignettes in
addition to cross-case thematic analysis (Padgett, Tran Smith, Henwood & Tiderington, 2012). The vignettes
of four participants (2 to 3 paragraphs in length) were supplemented by graphs, that is, horizontal lines with
call-out boxes showing the timing of traumatic events from birth to the time of the interview. For this report
with its identifying details (even with use of pseudonyms), we returned to participants and requested their
approval, which was generously given.

When seeking to illustrate a theme, the goal is to select the quotes with the best fit. If a direct quote contains
tangential information as well as key points, the use of ellipses (. . .) can indicate editing out the tangential
parts. Other than trimming, quotes should not be changed or language cleaned up. One other suggestion is to
link each quote with an ID number. In this way, reviewers and readers can see that different individuals are
being cited rather than one or two favorites.

Case vignettes feature description as well as illustration (Spalding & Phillips, 2007). Condensing and
describing a case—a person’s life or the inner workings of an organization or other entity—is a difficult task
since the study design entailed meticulous unpacking of the elements of the case for close examination.
Summarizing, graphing, and schematizing compresses a lot of information. The researcher exercises
responsibility in ensuring the presentation does not stray from the data along the way.

330
331
Writing Stance and Style

332
Rhetorical Devices Suited to Qualitative Studies
Successful qualitative reports give readers fresh perspectives and challenge them to think differently.
Qualitative research is tailor-made for rhetorical devices such as debunking, demystification, and irony.
Qualitative researchers are justifiably proud of their record of probing beneath carefully constructed facades
and coming up with counterintuitive findings. The religious leader who touts family values and privately
cheats on his wife, the mother who confesses to loathing her children as much as she loves them, and the
“lazy” men on the street corner who are really furloughed construction workers—all are stories that need to be
told. By doing so, qualitative researchers expose the incongruities and inequities taken for granted in social
life.

Using such rhetorical devices places the onus on the researcher to write in a convincing and credible way.
Being counterintuitive means asking the reader to change his way of thinking about something. This takes
compelling and intelligible writing. Carol Stack’s (1974) classic study of African American women living in
the Chicago projects struck home because she chronicled the resilience and resourcefulness of these women.
This portrait directly countered the “Black family as pathology” viewpoint among policymakers at the time
(and unfortunately continuing today). Similarly, Eric Klinenberg’s Heat Wave (2002) reports on Chicago’s
disastrous summer of 1995 as a slow-motion nightmare of public policy failures resulting in the deaths of 700
of the city’s most vulnerable and isolated citizens. This use of extensive documentation and interviewing to
explain the disturbingly high death toll can be seen as a form of abductive thinking.

Demystification is about clarification, normalizing what is poorly understood or hidden from view. Kranke and
colleagues interviewed adolescents diagnosed with a mental illness to jointly explore self-stigma and to
demystify mental illness (Kranke, Jackson, Taylor, Languth, & Floersch, 2015). Carmack’s (2014) study of
medical errors exposed how hospitals deal with such consequential yet misunderstood events in a doctor’s
work. Similarly, Metcalfe and Humphries (2002) used action research to demystify the experiences of being a
foster parent. Grypdonck (2006) argues that qualitative research is especially equipped to demystify the
evidence-based practice movement to “reduce it to its true proportions” (p. 1373).

Some qualitative researchers embrace subtle (and not so subtle) irony in framing the findings, thereby drawing
attention to gaps between what was observed and what is widely assumed (Lofland & Lofland, 1995). As
noted by Fine and Martin (1990), sociologist Erving Goffman (1961) was a master at using irony in Asylums.
Anthropologists revel in irony as a figure of speech used to expose inconsistencies and disconnects (Fernandez
& Huber, 2001). Of course, if taken too far or inappropriately, irony can appear mocking and condescending.
It is easier to dispense when leavened with humor.

Last but hardly least, qualitative reports feature the words and meanings of study participants. More often
than not, these individuals have had few opportunities to express themselves in their own words and be taken
seriously as an expert. And, when able to speak freely to an empathic listener, participants regularly call up
their own rhetorical devices and storytelling capabilities. This can be the most authentic and convincing
rhetorical device of all. One of the easiest arguments to make for conducting the NYSS and NYRS was to

333
urge that the opinions of formerly homeless mentally ill adults were worth listening to—and that such
accounts were largely missing from the voluminous literature on the problems of homelessness and serious
mental illness. The participants in both studies did not disappoint.

334
Writing Style: Rhythm and Grab
Successful qualitative reports have a rhythmic quality, weaving excerpts from the data into a seamless
exposition of the study’s interpretations. Such reports are confident, but not pompous, compelling but not
argumentative. Qualitative reports pique curiosity and encourage vicarious experiencing—they grab you
(Charmaz, 2006). Attention to reflexivity and self-awareness of the researcher-as-instrument lends an air of
humility and sincerity to the report (Tracy, 2010). There is still the necessary exposition of details on methods
and procedures—transparency demands no less. However, qualitative reports should be a good read.

Part of the grab of the qualitative report comes from the use of first-person pronouns. Another part comes
from the eloquence of study participants on display. Raw excerpts should push the narrative forward, not be a
diversion. McCracken (1988) likened this process to a small plane practicing takeoffs and landings. The plane
gains altitude (interpretation and exposition of study themes), but it also repeatedly touches down on the
landing strip (excerpting of raw data). The researcher aspires to flawless takeoffs and landings.

Jargon tends to narrow the audience to other professionals or academics and undermine the study’s relevance.
Professionals in a variety of fields have their own vernacular and idioms, whether it is arcane medical
terminology or the latest psychological argot. Qualitative researchers drawn to interpretivism are especially
adept at coining terms like prosthetic pedagogy, relational echoes, and performative diaspora (all drawn from titles
of articles in the journal Qualitative Inquiry). Artful terms can elevate meaning but they can also create
distance from study participants, policymakers, and practitioners. Gilgun and Abrams (2002) urge using
“concrete, everyday language” (p. 41).

335
Use of Metaphors and Other Tropes
Metaphors have been used to describe the qualitative research process (e.g., a “journey of discovery”) as a way
of capturing its excitement and uncertainty. One of the more enjoyable aspects of writing and reading
qualitative reports is the use of metaphors. Erving Goffman (1961) described the “careers” of mental patients
and Sandelowski and Jones (1995) told of the “healing fictions” that pregnant women devise to explain a
diagnosis of fetal anomalies. Other examples include Riessman’s (1990) divorcing couples that “mourn
different dreams” and Liebow’s (1993) portrayal of the “little murders of everyday life” endured by homeless
African American women. Respondents often use metaphors, whether commonplace (“walking on thin ice”)
or imaginative (the woman with breast cancer who says “I had a garden growing in my breast”).

Other tropes used to spice up a qualitative narrative include turning points and epiphanies (Denzin, 1989), each
recalling critical moments when events converge to alter a person’s life. Such figures of speech highlight what
is important and give a transcendent quality to the study and its interpretations. At the same time, describing
events as epiphanies or turning points presumes that participants share in this attribution (whether they use
the actual terminology or not).

336
Challenges and Rewards of Publishing
In an era of publish or perish, the academic researcher seeking career advancement has no choice but to
submit manuscripts for publication. However, dissemination should be the goal of all studies. For some
researchers, this goal can be elusive. They may apply unrealistic standards to their work and become paralyzed
by doubt. As Robert Weiss (1994) noted, “writing is exposure” (p. 205), and few exposures are more humbling
than submitting manuscripts for peer review. Although there has been improvement in recent years,
qualitative studies are still vulnerable to hostile or misinformed reviewers, editorial misunderstandings, space
limitations, and complaints about too little description of the methods. On a more upbeat note, journals are
open to qualitative studies as never before, and editors are mindful that they need to have qualitative
researchers carry out the review. The growth in qualitative-specialty journals has also been remarkable, leading
to diversity in topical area as well as approach (a list of these journals can be found at the end of Chapter 1).

Three basic formats exist for disseminating qualitative research: oral presentations, books, and journal articles.
Papers presented at professional meetings are typically peer reviewed, and acceptance rates vary with the
selectivity of the conference sponsor and abstract reviewers. Condensing a qualitative report into a brief
abstract for review (not to mention 15 or so minutes for the presentation) is daunting, but this format is the
staple of academic conferences. Poster presentations—a kind of passive oral presentation—are often easier to
get approved, but they are also a difficult fit since a poster is basically an extended abstract with visuals (tables,
graphs, etc.). Despite these drawbacks, oral presentations have value. They give the researcher an opportunity
to answer questions and to test out ideas with a live audience.

Qualitative reports are ideally suited for publication in book form because of their length—virtually all of the
classic works were books. Regrettably, the days when academic presses offered small press runs of scholarly
books are virtually over in this era of market-driven competition. Meanwhile, book publishers have turned to
peer reviews (in addition to the traditional standard of marketability) in determining the worthiness of a
manuscript for publication. Unlike journal manuscript submissions, a would-be book author may submit more
than one proposal at a time and await feedback from multiple publishers. Although each publishing company
has its own instructions, the following are common elements of what is expected: (1) a detailed overview of
the book with chapter titles and a table of contents; (2) biographical description of the author(s); and (3) a
description of the book’s projected audiences and the market for the book (including competing or
overlapping works). Qualitative studies do not grab a publisher’s interest unless there is a buying audience
readily identifiable. Thus, a compelling case needs to be made to the editor and the reviewers.

Publishing in peer-reviewed journals has all but displaced books as the preferred outlet in academia, much to
the chagrin of many qualitative researchers. Yet one need not despair, as journals are increasingly surpassing
their space limitations by offering online supplements and resources. Selecting the right journal may depend
on this particular option, but the decision should ultimately hinge on answering the audience question.
Academic journals tend to be identified with a discipline (such as social work or psychology), a field of
practice (such as mental health or child welfare), or methods (such as qualitative, mixed methods,
ethnography, and phenomenology). The first of these is more likely to favor authors from within their

337
respective discipline, but the last two are open to multidisciplinary submissions.

Double blind peer review is the norm, although some journals attach the authors’ names to the manuscript
(the reviewers are still unknown to the authors). Double blind reviews help prevent favoritism, but they do put
the onus on the authors to remove identifying information from the manuscript, including self-citations. They
do not, however, prevent the editor from exercising bias if prone to doing so. While most editors strive to be
impartial, some may exercise personal or professional prejudices in deciding whether to send the manuscript
out for review, whether to self-assign to be an additional reviewer, and in crafting the decision that goes with
the peer reviews. Split decisions among reviewers leave more discretion to the editor.

Choosing the right journal is best done in consultation with knowledgeable others. Some questions to ponder
include the following: the journal’s openness to qualitative reports and fairness of the reviews, the unevenness
of peer reviews, the wait time to receive the reviews (and/or to be published once accepted), whether the same
reviewers receive the revised manuscript, and so on. Some journals have successfully reduced reviewing times
but still have a long wait before an accepted manuscript gets published (online publishing has helped relieve
some of this backlog).

An additional suggestion for publishing is to develop a hierarchy of journals—an A list, B list, C list, and so
on. A-list journals usually have high impact, large audiences, and high rejection rates. Sometimes a
manuscript feels right for the A list—it has robust theoretical as well as empirical content. Other manuscripts
may have more modest goals and results (B or C list). A few years ago, I thought that we (the NYRS team)
had an A-list manuscript, but that assessment lost altitude as we were rejected by three A-list journals in
succession. The manuscript was revised and resubmitted to a less stringent journal for review.

A manuscript is essentially held hostage for as long as the journal takes to decide, so it is best to factor this
into the decision on where to submit. On a more upbeat note, journals have been increasingly transparent
about their processes of review and savvy about qualitative methods. A glance at the journal’s table of contents
can indicate if there is receptivity to qualitative reports.

The decision on when a journal manuscript is ready for submission is a difficult one, especially when
coauthors disagree. There are authors who need to feel the manuscript has undergone all possible iterations,
and there are those who take the middle ground of working to produce a publishable manuscript without
obsessing over it. I count myself among the middle grounders, and my coauthors are familiar with my
exasperated cries of “It’s time to let the reviewers decide!” Though I do not wish to submit work that is poorly
done, I also know that our best outcome from a journal review is to “revise and resubmit.” Since we anticipate
no lack of criticism, why do it all ourselves?

After consulting the journal’s author instructions, additional preparatory work is recommended. For example,
find out the names and affiliations of the editors and reviewers, become familiar with the journal’s content and
emulate it stylistically. Recall that the editor’s input can be instrumental, and he or she is not blinded to the
researcher’s identity, home institution, or so on.

The overriding message with regard to publishing is one of tenacity and hope. I like to think that every

338
worthwhile study has a place somewhere in the universe of options. Journals nowadays have impact factor
scores, and rejection rates available to would-be submitters so that one can compile a list of potential outlets
based on degree of difficulty as well as fit. Unless explicitly told not to resubmit, I consider rejection letters
accompanied by reviewers’ comments as invitations to try again. Assuming the suggested revisions are
reasonable, I faithfully make them and resubmit. If a suggestion appears unreasonable, I politely counter argue
that it does not work for the study. Revisions may have to go a second or even third round, but take heart that
the editor usually wants you to succeed after investing so much effort mediating between you and those picky
reviewers. Even when a manuscript is rejected outright, reviewers’ comments offer helpful tips for revising and
resubmitting it elsewhere.

As difficult as it may seem at times, a willingness to roll with the punches is the best survival mechanism. This
is especially important advice for researchers early in their careers. In recent years, I have seen response times
improving both as a submitter and a peer reviewer. The pressure from journals to complete the review—with
deadlines as early as two weeks after an agreement to review the manuscript—is no doubt a response to
unhappy authors. As a reviewer, I typically have a month to complete the review (with editors willing to give
extra time if asked). As an author, I give a journal 60 to 90 days before inquiring about the manuscript’s fate. I
do not hesitate to politely ask (and reask) if necessary.

For many qualitative researchers, squeezing findings into 25 or fewer pages is a lot to ask. For most, however,
the prestige and career enhancement make it worth the effort. Even qualitative-friendly journals and reviewers
must grapple with the lack of consensus on standards for quality and rigor. Submitting a manuscript for
review can feel like trying to hit a moving target (or, worse, one hidden from view). I have experienced one set
of journal reviewers complaining about too many direct quotes without interpretation and another set
critiquing the amount of interpretation vis-à-vis direct quotes. I have also had grounded theory experts express
dissatisfaction that a fully developed grounded theory model was not forthcoming, even though the majority
of grounded theory studies do not reach this pinnacle nor is it always appropriate. As discussed in Chapter 8,
this unevenness is a natural consequence of the lack of consensus on standards. Nevertheless, the situation has
improved considerably in recent years as journal editors have become sensitized to qualitative methods.

The way that impact scores are calculated disadvantages newer journals having fewer articles in their archives
to be cited. Being newer on the scene, qualitative research journals have lower impact scores. The journal
Qualitative Health Research—one of the oldest—has an impact factor score of 2.1 compared to the New
England Journal of Medicine score—a whopping 55. The increasing emphasis on rankings affects scholars in
the practice professions but it is usually taken with a grain of salt by even the most metric-centric researchers
(Moustafa, 2015).

A final note on dissemination: The growth in online publishing has made qualitative research more widely
accessible. Noteworthy examples include The Qualitative Report, Forum: Qualitative Social Research (Forum:
Qualitative Sozialforschung), and the International Journal of Qualitative Methods. Electronic access will
undoubtedly grow in the future, although the challenges of providing rigorous peer reviews will remain.

339
340
Criteria for Review: The Qualitative Dilemma
As discussed in Chapter 8, we have no paucity of guidelines intended to specify what is high-quality
qualitative research. Some of these are directed to researchers and intended to assist them in crafting the study
as well as writing it up for publication. Others are designed for journal reviewers to consult to improve the
quality (or at least the fairness and reliability) of manuscript reviews.

The disciplinary foundation and mission of the journal is an important aspect of the review process. In
nursing, qualitative research has attained greater parity with quantitative research. In medicine and public
health, qualitative studies have appeared more often in recent years, but few journals offer guidance. Sociology
and anthropology have discipline-specific norms with an emphasis on theoretical frameworks and critical
theorizing as well as methodological expectations. With an overwhelmingly quantitative foundation,
psychology is relatively new to qualitative methods but has nonetheless emphasized its own tradition of
phenomenology and narrative work (Gergen, Josselson, & Freeman 2015). Social work research is eclectic,
with most social work journals welcoming qualitative research and a dedicated journal (Qualitative Social
Work).

Recent efforts to consolidate and promulgate qualitative guidelines from the vast domain of health-related
research have resulted in an alphabet soup of acronyms: COREQ, RATS, and SRQR. The Consolidated
Criteria for Reporting Qualitative Research (COREQ) is comprised of 32 items derived from an exhaustive
search of journal criteria, meta-syntheses, and existing measures (Tong, Sainsbury, & Craig, 2007). The
unfortunate acronym RATS (Relevance of the question, Appropriateness of the qualitative methods,
Transparency of procedures, and Soundness of the interpretation) is comprised of 25 questions (Clark, 2003).
The Standards for Reporting Qualitative Research (SRQR) is a 21-item checklist based upon reviewing the
literature and consulting with expert methodologists (O’Brien, Harris, Beckman, Reed, & Cook, 2014).

While still largely confined to the conduct of meta-syntheses, these guidelines are valiant attempts to foster
transparency, consistency, and explicitness. Here are ten criteria common to the checklists that are less likely
to cause controversy:

1. The research questions have convincing relevance and significance for theory and/or practice.
2. How study participants (or cases) were selected and recruited into the study is described.
3. Institutional review board (IRB) review and attention to ethical issues including confidentiality and
sensitivity is assured.
4. Researcher characteristics and reflexivity are described.
5. The research setting and context are described.
6. Data collection is described (types of data; use of audio-recorders; interview guides; field note guides;
transcription procedures; multiple vs. single interviews; use of QDA software).
7. Data analysis methods are described and consistent with the approach.
8. Interpretation procedures are described such that themes or other findings are understandable and
linked to the data.

341
 9. Use of quotes or other forms of raw data are used to illustrate themes or other findings.
10. Findings (including serendipitous findings) are tied to the original research questions as well as to the
larger knowledge base.

Overly detailed criteria—especially a few of the COREQ’s 32 items—can appear intrusive (e.g., stating the
researchers’ occupation at the time of the study and if nonparticipants were present during interviews) or
oddly idiosyncratic in use of terms (description of the “coding tree”). Although still largely confined to meta-
syntheses, criteria for quality are not likely to go away anytime soon. The popularity of meta-analyses and
meta-syntheses—and a concomitant desire to include qualitative studies in efforts to identify best practices—
prioritizes the need for a standard metric for assessing quality. For example, Leamy and colleagues used the
RATS in their narrative synthesis of the literature on personal recovery in serious mental illness (Leamy, Bird,
LeBoutillier, Williams, & Slade, 2011). Posing and answering the questions through independent ratings of
97 studies, the authors tested their quality scores for concordance and found agreement to be high.

Here are two final points on standards-based decisions in publishing and qualitative research. First, if there is
disagreement on what the criteria should be, there is growing consensus about what they should not be. Gone
(almost gone or should be gone) are the days of demanding hypotheses, large samples, generalizability,
interrater reliability, and replicability. Second, applying criteria to a qualitative report evaluates not just the
quality of the study but how well or poorly it was written up—a rigorously conducted study with a poor write-
up will reap negative reviews from peer reviewers. Whether one agrees with preset criteria or prefers a less
standardized approach, the quality and probity of the writing is undeniably important.

342
Summary and Concluding Thoughts
This chapter has offered a number of suggestions for writing the qualitative report, all proffered in the spirit of
“take what is useful and applicable for your needs.” Much depends on the structure, format, and style of the
writing. The decisions start early and pivot around questions of audience, researcher reflexivity, use of
numbers, and authorship responsibilities. Most write-ups follow the conventional outline (introduction,
methods, findings, and conclusions) but can vary depending on which qualitative approach is used and the
outlet for dissemination (journal, book chapter, report, thesis, etc.).

The researcher is no longer invisible in the qualitative report, but the operative word here is trust. Not blind
trust but a form of earned respect based upon candor without overdoing disclosure. Readers have a vested
interest in knowing about the instrument. Weaving in this information—or sidelining it in an appendix or
footnote—is a stylistic decision that is made with care.

A qualitative report tells a story. In a rhythmic fashion, it blends description and interpretation. To enliven
the argument and make it more compelling, qualitative writers use a number of rhetorical devices including
metaphors, irony, and other tropes. The writing should have minimal jargon and be understandable to a broad
audience. Intense scholarly works have to employ a higher level of theoretical and conceptual discourse, but
they need not be obscure and off-putting. Successful qualitative studies pay heed to rigor, but they are also a
good read, spirited and thought-provoking.

The researcher’s capacity to be surprised and to present serendipitous findings in an engaging manner can
make the difference between a ho-hum report and one that grabs an audience and stays with them. Every
well-conducted qualitative study has a home out there, whether it is in a top-tier journal or one farther down
the prestige hierarchy. Without dissemination, all of the hard work that came before and all of the cumulative
growth of knowledge that can come after is imperiled.

343
Exercises
1. Go to one or more of the qualitative journals mentioned in this chapter and in Chapter 1. Using key
word search techniques, select an article of interest to you that represents one of the qualitative
approaches and bring the article to class for discussion.
2. In class discussion, present your selected article, highlighting the style of writing used in the article. Do
the authors use metaphors or other narrative techniques to get their point across? Do they use tables or
figures? If so, how?
3. Locate the journal’s guidelines for authors submitting manuscripts. Do they include adequate
description regarding what the journal expects?
4. From the list of monographs at the end of Chapter 1, check one out from the library, and answer the
following questions. Is there a chapter or appendix on methods? Does the author talk about his or her
role in the study? If so, how much detail is offered? Are figures, tables, and/or photographs used? If so,
how do these visual displays help tell the story?

344
Additional Readings
Becker, H. (2007). Writing for social scientists (2nd ed.). Chicago, IL: University of Chicago Press.

Gilgun, J. F. (2005). “Grab” and good science: Writing up the results of qualitative research. Qualitative
Health Research, 15, 256–262.

Holliday, A. (2007). Doing and writing qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Janesick, V. J. (2015). “Stretching” exercises for qualitative researchers (4th ed.). Thousand Oaks, CA: Sage.

Padgett D. K. (2004). Spreading the word: Writing and disseminating qualitative studies. In D. K. Padgett
(Ed.), The qualitative research experience (pp. 285–295), Belmont, CA: Thomson.

Prendergast, C. (2004). The typical outline of an ethnographic research publication. Teaching Sociology, 32(3),
322–327.

van Manen, M. (2006). Writing qualitatively, or the demands of writing. Qualitative Health Research, 16,
713–722.

345
10 Mixed Methods

Mixed methods represent a new wave in research, a “third paradigm” according to Johnson & Onwuegbuzie
(2004), and their popularity is a tribute to the growing influence of qualitative methods (Hesse-Biber, 2015;
Morgan, 2014; Tashakkori & Creswell, 2007). Some research topics (the prevalence of a social problem) are
manifestly quantitative; others such as gang initiation rites are undoubtedly qualitative. In between lie myriad
opportunities to use both approaches for synergistic ends. Among the contributions of qualitative methods to
mixed methods are their ecological validity and engaging accessibility.

To be sure, researchers have been mixing methods for a long time. Pelto (2015) points out that ethnographers
have long used psychological tests as well as measures of caloric intake, social network size, and so on. Urban
sociologists have relied on surveys and immigration data as well as neighborhood ethnography. The difference
lies in mixed methods having reached a tipping point in popularity in the past few years such that the pursuit
of mixed methods has taken on a life of its own.

As a consequence, the literature on mixed methods has been expanding rapidly. Since the launch of the
Journal of Mixed Methods Research in 2007, the number of books about mixed methods has grown from a
handful to 30 and counting (Onwuegbuzie, 2012). Also telling was a panel convened at the National
Institutes of Health in 2011 to develop criteria for high-quality mixed methods grant proposals (a panel on
which yours truly participated). The reason for organizing the panel was practical: Several of the Institutes of
Health were dealing with rising numbers of mixed methods proposals, and grant reviewers had little guidance
on what constituted a high quality mixed methods study. The guidelines that resulted were dedicated to
helping researchers write successful proposals (Creswell, Klassen, Plano Clark, & Smith, 2011).

All of this attention has brought elucidation, but it has also highlighted plenty of gray areas, especially when
mixed methods studies have complicated (and convoluted) designs. Variation is inevitable given the many
options available, and learning about mixed methods can be daunting given so many out-of-the-box
applications (Bazeley, 2010). This chapter summarizes what is known and describes both the challenges and
rewards of mixing methods.

346
The Epistemology Question in Mixed Methods
It is fair to say that the “third paradigm” has not been universally accepted within qualitative methods. Some
qualitative researchers dismiss mixed methods as ‘‘positivism dressed in drag’’ (Giddings, 2006, p. 195). For
their part, quantitative researchers are rarely concerned with epistemological objections to mixing methods,
but such mixing preoccupies a subset of qualitative researchers for whom the “incommensurability” thesis
holds true (Denzin & Lincoln, 2011). In their book The Constructivist Credo, Lincoln and Guba (2013) argue
that postpositivist and constructivist paradigms cannot be integrated. Thus, mixing methods is not plausible.
While quantitative and qualitative methods might become commensurable under a future “meta-paradigm,”
such a unifying framework has yet to be developed (Lincoln & Guba, 2013, p. 60).

Their popularity a tacit nod to pragmatism, mixed methods have garnered attention in some of the most
quantitative areas of research, for example, randomized clinical trials. This popularity has made one concern
vital for all qualitative researchers: ensuring that the qualitative side is given its due. Enthusiasm for mixed
methods has had a breakthrough moment largely because the dominant quantitative side came around.
Despite “ecumenical rhetoric” coming from quantitative researchers (Katz & Mishler, 2003, p. 49), the
methodological heavy lifting in explicating mixed methods has come from qualitative researchers. This may be
true by default (the existence of a smaller and more dedicated community on the qualitative side), but it is also
a consequence of genuine concern that mixed methods could too easily devolve into superficiality. Working
with a very powerful partner can be beneficial, but it can also reduce the less powerful side to a shadow of its
former self.

347
The Rise of Mixed Methods and Their Rationale(s)
Appreciation for mixed methods came in part from growing recognition that existing methods are not up to
the task of solving “wicked problems” involving multiple interacting systems, for example, climate change,
human trafficking, and so forth (Mertens, 2015). Systems science and sophisticated statistical modeling were
developed to bring in larger contexts and multilevel units of analysis (Diez-Roux, 1998), but their utility
depends on having access to large-scale datasets generated from complex social and economic systems. Even
in the rare occasion where such data are available, there are inevitably missing parts.

In this context, mixed methods designs combine the granularity of quantitative data with the naturalistic
holism of qualitative data. There are many problems that can be studied using this one combination—sex
trafficking, natural disaster relief, gender violence, and environmental accidents. On a less ambitious scale—
and this is obviously most often the case—mixed method studies focus on topics amenable to completion by
smaller-scale collaborations (Bryman, 2006; Creswell & Plano Clark, 2011; Tashakkori & Teddlie, 2010).

In their seminal article, Greene and colleagues identified five reasons for carrying out mixed methods studies:
triangulation, complementarity, development, initiation, and expansion (Greene, Caracelli, & Graham, 1989).
Triangulation, the earliest and most widely invoked of rationales, refers to comparisons across the methods to
reveal corroboration or discrepancies (Morse, 1991). As discussed in Chapter 8, the presumption of a fixed
point of reference or truth waiting to be located is problematic for many qualitative researchers (Flick, 2004;
Sandelowski, 2000). Complementarity refers to enhancement or clarification. Thus, the two sides represent
different pieces of the puzzle. Development takes place when the two sides happen in sequence, and one
influences the other temporally. Initiation occurs when findings conflict and new frameworks are developed.
Expansion refers to the broadened theoretical understanding that comes from juxtaposing qualitative and
quantitative perspectives. These archetypal categories represent possibilities, a starting point for thinking
about why one might mix methods in the first place.

348
Integrating Mixed Methods
The most immediate and least understood challenge of mixing methods is that of integration (Bryman, 2006),
or what some call the 1+1=3 mandate (Fetters & Freshwater, 2015). Creswell and Plano Clark (2011) propose
three ways to integrate in mixed methods studies: merging, connecting, and embedding. Merging occurs when
the two sides and their findings are juxtaposed or interwoven. A hypothetical example might be a qualitative
typology of racially biased attitudes compared to participants’ scores on a measure of racial bias. Connecting
implies that one side informs the other side in a temporal sequence. Probably the most common type of
integration, this allows each side to hold its own without interference since they are separated by time. Thus,
analyses of focus group data might be used to inform the development of a survey questionnaire or a measure.
Likewise, the results of a survey might lead to purposive sampling of participants for follow-up individual
interviews. Embedding or nesting takes place when one side is subordinated to the other and used in an
ancillary way. An example of this is the use of interviews, observation, or document review (or all three)
during a large randomized controlled trial (RCT) to provide in vivo information on how the trial is
proceeding and to independently assess participants’ progress.

349
Types of Mixed Methods Designs
All research designs operate from a premise of intentionality, especially mixed methods designs where the
optimal goal is to link or integrate. The shorthand way of designating the relationships between the QUAL
and QUAN sides is a directional arrow for sequential designs (e.g., QUAL->QUAN) and a plus sign for
concurrent designs (e.g., QUAL+QUAN). To connote subordination, lowercase can be used. Thus a qual-
>QUAN design refers to a study that begins with a small qualitative component that leads to a large
quantitative component. These shorthand designations are most useful for depicting uncomplicated mixed
methods designs (Guest, 2012).

The challenges of establishing definitional boundaries and limning the possibilities of mixed methods are
twofold. First comes the lack of solid consensus on what constitutes mixed methods. Is a large survey with a
few open-ended questions mixed? Is a qualitative interview study that includes a standardized measure mixed?
Is a qualitative study conducted long after a quantitative study has ended (or vice versa) part of a mixed
methods study? I would say no to all of these (at best, they are weak mixed methods studies). In the first
example, the qualitative information—coming amidst or after crisp Likert-type responses—will likely be thin.
With the second example, a numeric measure of something might have descriptive value (especially if part of a
case study or ethnography where other types of data are collected), but it cannot offer much more. The third
example violates the intentionality and integration premises of mixed methods. The post-hoc study may be a
sequel, but the passage of time makes integration much less viable.

The second challenge arises from the fact that complicated mixed methods designs defy easy typologies
(Guest, 2012). Their components—with feedback loops that can redirect a study toward new sampling or data
collection—are often easier to diagram after the fact than to label in advance (Mendlinger & Cwikel, 2008).
Even when diagrammed, complicated mixed methods studies frequently need to be narratively described to
convey their twists and turns. In the following sections, we begin with the basic options then move toward
more complicated designs and use examples along the way.

350
Sequential Designs
Sequential mixed method designs are probably easiest to understand. Termed connecting integration in the
NIH Guidelines (Creswell et al., 2011), the points of contact in these studies come from linking an ending to
a beginning, leaving the processes within each study intact. Wackerbarth, Streams, and Smith (2002), for
example, used individual and focus group interviews to generate items for a survey of Alzheimer’s’ caregivers.
Positioning the qualitative study first is the most common variant of sequencing, and instrument development
studies are common in the literature. The example given in the Creswell et al. (2011) NIH Guidelines was of
a mixed method study of alcohol drinkers with Hepatitis C (Stoller et al., 2009). Using interviews and
Internet data from blogs and postings, the investigators developed items for a measure of the drinking
behaviors that affect “Hep C” status and administered the measure as part of a questionnaire.

The reverse order—a qualitative study coming after a quantitative study—would be exemplified by a
structured interview study from which a purposive subsample is drawn for in-depth interviews. In 2001, I
became involved in the post-hoc qualitative portion of the New York Housing Study, a quantitative
randomized experiment of 4 years’ duration (Tsemberis, Gulcur, & Nakae, 2004). In its final year of data
collection, the study’s design became QUAN->qual as prior statistical analyses revealed a puzzling lack of
group differences on several key outcomes (Padgett, Gulcur, & Tsemberis, 2006). The “small qual” segment
of the study entailed appending a qualitative interview to the final structured interview. These questions were
audiotaped and analyzed for patterns and themes that could shed light on the earlier quantitative findings.

Paul Farmer’s ground-breaking work with AIDS patients in Haiti has been cited as an illustration of how
sequential mixed methods can be synergized to yield deeper understandings and theoretical framings (Katz &
Mishler, 2003). Trained as an anthropologist and physician, Farmer (1999) was initially interested in
conducting ethnography to understand the impact of the disease on patients and their families. This work
raised new questions that could only be answered through surveys and epidemiological methods. Over time,
the pervasive influence of historical and structural factors was brought in to understand the ways that Haitian
families and communities coped with the rising tide of AIDS infections. Although it is not apparent that
Farmer intended to carry out mixed methods (and such terminology and self labeling were rare in the early
1990s), Katz and Mishler (2003) use his sequential approach to illustrate fairness in mixed methods, that is,
Farmer treated both sides as equally valuable, neither positioned as the standard for validating the other.

351
Concurrent Designs
In concurrent designs, one method may be dominant over the other (QUAN+qual or QUAL+quan), or they
may be given equal weight (QUAN+QUAL). In the NIH Guidelines, the forms of integration associated
with these designs are merging and embedding (the latter connoting a “dominant-less dominant” design). Box
10.1 offers an example of a QUAN+qual study carried out in different nations.

352
 Box 10.1 A Mixed Method Study (QUAN+qual) Testing a Measure of
Social Capital in Peru and Vietnam
Mixed methods have an intrinsic appeal for instrument development and testing because most measures’ underlying constructs are
complex and open to differing meanings and interpretations. One such concept, that of social capital, has become widely used as an
indication of the ways that social relationships may confer health benefits, from fostering a sense of belonging to providing links to
valuable resources. The measurement of social capital at the individual level is seen as a potential indicator of health in general and
access to health care in particular (Szreter & Woolcock, 2004).

DeSilva and colleagues (2006) developed a measure of social capital (the SASCAT), translated it into Spanish and Vietnamese and
administered it to a large sample of children’s caregivers (3,000 in Peru and 2,771 in Vietnam). In addition to psychometric tests of
the measure’s validity, the researchers criterion-sampled 20 Peruvian and 24 Vietnamese respondents for in-depth interviews. These
“cognitive interviews,” lasting from 1 to 2 hours, elicited further thoughts and ideas related to each SASCAT item. An example of
an item is “In the past 12 months have you joined together with other community members to address a common problem or issue?”
The interviews were audiotaped and content analyzed to see if (and how often) open-ended comments diverged from the authors’
original intention regarding each item’s meaning.

The findings were revealing. Although the quantitative factor analysis results from the two countries were strikingly similar, the
qualitative interviews brought several cultural misunderstandings to the surface. The concept of “community,” for example, was
readily accepted in Vietnam but not understood by many Peruvians (who defined it as one’s social support network, not the
surrounding area). In both Peru and Vietnam, “trust” was not considered something one can impute to the “community” in general
but only to known individuals. Similarly, “help from others” was largely defined as economic support—contrary to the measure’s
inclusion of emotional support within the definition. The authors understandably concluded that cognitive validation needs to
precede instrument development (DeSilva et al., 2006).

Commentary: In this study’s design, the two methods were used for corroboration as well as completeness (i.e., the researchers did
not posit a single meaning for each item but instead sought out multiple meanings to improve the measure). The findings
demonstrate the importance of qualitative methods in cross-cultural research in which subjective meanings can vary along cultural
lines.

In QUAN+qual designs, researchers may collect qualitative data to enliven or illustrate the quantitative
findings, for example, excerpts from responses to open-ended questions or case vignettes (Morgan, 1997). In
the reverse QUAL+quan approach, qualitative researchers might collect quantitative data via standardized
measures, or they might use supplementary quantitative data from documents or archives. Snow and
Anderson (1991) made use of tracking data from various agencies to supplement their intensive interviews and
ethnographic observation of homeless men. The resulting depictions contained both statistical and
ethnographic description.

Using quantitative data can be risky with small samples, but if done judiciously it need not detract from the
inductive, emergent nature of a qualitative study. Similarly, the inclusion of ancillary qualitative data does not
challenge the primacy of a “big QUAN” study. A fully integrated QUAL+QUAN study is probably the least
common mixed methods type due to aforementioned demands on time and resources.

353
Complicated Designs
Once we leave the accessible terrain of simple mixed methods designs, we encounter a plethora of possibilities.
This is especially true when the study happens over a period of time to allow points of contact to be worked
out in different ways. One thinks of oscillation or a spiral as one graphic representation (others might view
these designs as convoluted). Mendlinger and Cwikel (2008) chose the DNA double helix as an analogy for
the mixed methods design in their study of immigrant mothers in Israel. Thus, two parallel spirals with
bridges between them—much like the rungs of a twisting ladder—formed the architecture of the study in
which seven turns of the helix marked its stages. The multiple forms of data collection (ethnography, focus
groups, individual interviews, and a survey) created a foundation from which sequential as well as embedded
mixing took place. Thus, focus groups about why women did not breastfeed yielded multiple items that were
subjected to factor analysis. In the study’s final phase, a large-scale telephone survey was conducted to learn
more about women’s knowledge of and attitudes about menstruation (Mendlinger & Cwikel, 2008). Spiraling
around like this is labor intensive and assumes the researchers can maintain flexibility as well as fidelity to the
demands of the design.

354
Box 10.2 Strange Bedfellows? Embedding Qualitative Methods within
a Randomized Controlled Trial (RCT)
Kaptchuk and colleagues (2009) intensively interviewed a subset of 12 of the 262 patients enrolled in a 6-week single blind RCT
testing the role of practitioner attentiveness (less vs. more) and acupuncture treatment (placebo vs. real) in relieving symptoms of
irritable bowel syndrome (IBS). IBS was an ideal condition for examining placebo effects as its severity can only be assessed through
patient self-reports. The design of the RCT afforded the opportunity to interview patients about their expectations, hopes, and
doubts during the trial. Was the placebo viewed as working? How did patients experience treatment that may not have been real? All
of the study participants experienced improvement in IBS symptoms (and several reported other benefits such as improved social
relationships). The qualitative findings challenged psychological theories of expectancy effects by showing that patients interpreted
their responses in differing ways over time.

In another example of strange bedfellows, Plano Clark and colleagues (2013) embedded an interpretive qualitative approach within
an RCT testing a treatment for cancer pain. An earlier test of the experimental treatment (a psycho-educational self-management
approach) showed positive effects in the aggregate compared to treatment as usual, yet half of the experimental group did not
improve significantly in reducing their pain. The qualitative data showed that the 6-week duration of the intervention was too short
and the nurse-patient interactions too limited. The second RCT offered an expanded version of the treatment (low dose vs. high
dose). In this trial, qualitative aims explored difficulties, positive strategies, and nurse-patient interactions over the course of the trial.
Although envisioning the qualitative interviews as following a template approach, the researchers realized that the questions called
for an interpretive qualitative approach to explore subjective, contextualized, and socially constructed meanings of patients’
experiences.

Commentary: The Plano Clark et al. (2013) study demonstrates that not all interpretivist researchers agree with the
incommensurability thesis. Although at first glance an odd coupling, an RCT is often large-scale enough and long enough to
accommodate an agile junior partner as was done in this study. The success of method integration in these instances depends on the
responsiveness of the RCT protocols to the flexibility and additional time required by qualitative research. The benefits of timely
feedback derived from in-depth subjective information complemented and expanded the findings of the RCT and made it more
effective.

355
Mixed Methods: Ways of Going About It

356
Structural and Design Decisions: What, When, and How?
Making decisions about carrying out a mixing methods study requires that we break down the research
process and decide which phases will (or should) have points of contact and which will not. Are there
constraints on doing this, or can one mix and match at will? Consider the following series of pragmatic
statements that might help answer this question:

1. Methodological approaches are not inherently tied to specific paradigms or epistemologies (Pelto,
2015). For example, grounded theory has both postpositivist and constructivist versions.
2. Data-collection techniques (interviews, questionnaires, observation) are not inherently tied to
methodological approaches. Most qualitative approaches are open to using any or all types of data.
3. Data analyses are not inherently tied to techniques of data collection. It is hypothetically possible to
“quantitize” qualitative data or to “qualitize” quantitative data (Sandelowski, 2000).

What is the point of these statements? To emphasize that the constraints are relatively few, and the options
are many in deciding the “what, when and how.” With regard to the first statement, most mixing takes place
below the paradigm level (Morgan, 2014), and this is primarily a matter of being paradigm-agnostic as
opposed to paradigm-atheistic (meaning most qualitative researchers ignore the need to cite a paradigmatic
allegiance rather than disavow the importance of paradigms). The second statement points to the mixing of
methods through incorporating one or more qualitative data collection modalities. The third statement is a
reference to data transformation during the analysis stage (see Box 10.3 for an example). According to
Sandelowski (2000), most mixing takes place “on the shop floor of research” (p. 246) during sampling, data
collection, and data analysis.

Moving through these three orienting statements, we see the capacity to mix and match quantitative and
qualitative methods at different stages of a study. To be sure, some combinations do not work, for example,
narrative analysis and a survey are a poor fit. One should not mix and match without considerations of
appropriateness. Allegiance to the respective sides means that sampling is likely to be probability and
purposive, data are both quantitative and qualitative, and data analyses are both statistical and thematic or
narrative.

Sequential designs leave open the opportunity for each side to remain intact (assuming a reasonable
connection is made). In concurrent designs, the parallel processes, or “strands” (Tashakkori & Creswell, 2007,
p. 3) may intersect at one or more phases as in the double helix mentioned earlier. The lowest level of
concurrent mixing intensity is when the two sides stay separate and come together only at the end when
findings are juxtaposed. This barely reaches the threshold of integration.

357
 Box 10.3 “Quantitizing” Data in a Grounded Theory Study of Breast
and Prostate Cancer Online Discussion Boards
Online chat rooms and discussion boards offer an abundance of narrative data for qualitative analysis. Gooden and Winefield (2007)
used grounded theory and a “quasi-numeric” (p. 103) approach to examine gender differences in language styles and communication
among cancer survivors communicating online. They started with a hypothesis positing greater use of emotional communications by
women and greater use of informational communications by men. They examined online communications among 69 women with
breast cancer and 77 men with prostate cancer by using open, axial, and selective coding conducted independently by two readers.
The number of codes per message (or posting) and the frequency with which individuals posted were calculated and displayed in
tables in the published article.

From these analyses, two selective codes (“information support” and “emotional support”) were identified along with their respective
axial and open codes. Examples of axial codes included “facts about the disease” (under information support) and “coping
philosophies” (under emotional support). Instances of open codes were counted in each database and categorized proportionately
under each of these two main headings. As a result, Gooden and Winefield (2007) found that information communication
comprised 60% of women’s communications and 64% of men’s communications. Thus, there were few gender differences in the
frequency of emotional (versus informational) communication. Virtually all of the results section of the article was devoted to
describing the codes, thereby revealing subtle but meaningful aspects of gender. Under “information support,” for example, men were
likely to offer detailed factual information compared to briefer informational summaries supplied by women. Under “emotional
support,” women used warm dialog and affectionate phrasing while men suggested to their peers that they “keep their chin up” and
“beat the bastard.” In other code domains, such as use of humor and group spirit, men and women did not differ.

Commentary: This study’s use of a hypothesis and a QUAL+quan design set the stage for the quantification that followed. However,
the quantitative findings comparing men and women were modest and anticlimactic. In the study’s write-up, the numbers told a
small story, but the qualitative themes and interpretations were the main event.

With regard to the “what” question, Bryman (2006) reports that the concurrent mixing of standardized
surveys and qualitative interviews is a popular combination, the latter often based on a purposively selected
subsample from the larger survey sample. From the qualitative side, focus groups are a popular choice but with
them come the risk of superficial data (Agar & McDonald, 1995). In-depth interviews and in situ observation
are other viable options.

358
Incorporating Existing Measures Into Qualitative Interviews
There are a plethora of measures available for all manner of cognitive, emotional, and behavioral phenomena.
Miles and Huberman (1994) suggest that checklists or measures be used when the domain of interest is
measurable and the researcher wishes to make standardized comparisons across cases. The decision to do this
is not without consequences. First, standardized measures detract from what is otherwise a more free-flowing
experience in data collection. When used, they should be reserved for the interview’s end or a follow-up
interview. Second, their quantitative properties introduce numbers into the analyses, even though sample sizes
are rarely large enough to sustain statistics beyond the descriptive level of frequencies, ranges, and averages.

359
Quantifying Qualitative Data
As a researcher who treasures qualitative data, it verges on sacrilege to contemplate compressing or contorting
such data into variables. And yet I have done it and found it worthwhile for very specific purposes. Like so
many other aspects of qualitative inquiry, there are degrees of desecration (that is, transformation). The most
transforming and least satisfying scenario treats the qualitative data solely as raw material for variable
production. Content analysis of documents typically comes up with counts and frequencies, but this seems
appropriate when the documents lack depth (newspapers, meeting minutes, Internet blogs, etc.). Cluster
analysis, a statistical technique to identify typologies, has been applied to qualitative data (Macia, 2015), but
this also deprives the data of nuance and deeper meaning.

Of course, all qualitative data analyses must reduce masses of raw data in some meaningful yet forceful way.
The pertinent question is whether such reduction honors the depth and complexity of the data. In an effort to
have it both ways—an attempt made possible by having relatively standardized in-depth information collected
over time—both the New York Services Study (NYSS) and the New York Recovery Study (NYRS) entailed
quantifying qualitative data as an analytic option. Neither of these analyses supplanted the overwhelmingly
qualitative focus of their respective studies and findings, but they served a purpose nonetheless.

In the NYSS, we were stirred to quantify by witnessing a spirited debate in the research literature about
whether the “housing first” (HF) approach resulted in reduced dependence on drugs and alcohol. Rigorous
study findings had established that formerly homeless individuals with serious mental illness and co-occurring
substance abuse could move into their own apartment and remain housing stable for years (Tsemberis et al.,
2004). What interested us was the absence of positive effects on use of drugs and alcohol when comparing HF
clients to usual care persons living in more restrictive congregate settings (Padgett et al., 2006). Having three
in-depth interviews conducted over 12 months in the NYSS, we reasoned that we had substantial ecological
validity to ascertain participants’ substance use across the 12 months (including brief monthly interviews via
telephone).

Taking all we knew about each participant, we consensually assigned a 0 (for no illicit drug use and very rare
alcohol use) or 1 (for any illicit drug use and moderate to heavy alcohol use) to each person then conducted a
logistic regression analysis using this binary as the dependent variable (Padgett, Stanhope, Henwood &
Stefancic, 2011). With a sample size of 75, we were able to include predictors of race, gender, and baseline
substance use as well as program type. The results were striking: HF participants were three times less likely to
use substances than the comparison group (Padgett et al., 2011). This use of quantified qualitative data to
address an important policy question about the effects of HF was not part of our study’s original aims, but it
was a serendipitous direction made possible by the breadth and depth of the data.

Statistical tests with small sample sizes usually involve nonparametric tests building on cross-tabulations (Chi
Square tests) and simple group comparisons (t tests). For example, the Fisher’s exact test is appropriate for
testing the relationship between two dichotomous variables with a small sample size. Since multivariate
analyses require adequate numbers, a mixed methods design dependent on these would need to recruit a large

360
sample and adequate statistical power to detect relationships if they exist (see Box 10.4 for an example).

Figure 10.1 Mixed Methods Using Quantified Qualitative Datain the NYRS

361
Box 10.4 Quantifying Qualitative Data in the NYRS: Measuring
Trajectories of Recovery
The NYRS was designed as an all-qualitative longitudinal study of mental health recovery among newly housed psychiatric
consumers who had histories of homelessness and substance abuse. We reckoned that multiple interviews and ethnographic
observation over 18 months would be optimal in examining recovery from the insider perspective. Multiple domains of recovery
constituted the infrastructure of data collection, including mental status, physical health, use of drugs/alcohol, family and partner
relationships, jobs and income, housing satisfaction, and meaningful activities or hobbies.

After data collection ended, we were happily inundated with qualitative data to conduct case studies of 38 participants’ lives after
they had finally obtained housing stability. In addition to our qualitative analyses, we decided to quantify some of the data to make
calibrated comparisons of recovery trajectories across cases and over time (18 months). Participants’ interview transcripts for each of
four waves of data collection (at 0, 6, 12, and 18 months) were read by two study team members and rated 1 to 3 for low, mixed, or
high recovery based upon the eight recovery domains mentioned above. Consensus was used for discrepant ratings to produce a
recovery score from 8 to 24 for each person at each wave.

Ratings were graphed and plotted to depict trend lines, and a criterion was set: one standard deviation difference in the baseline
versus 18-month recovery score was considered significant change. Of the 38, only 15 showed change using this criterion: seven
negative trajectories and eight positive trajectories. Examining the qualitative data for these 15 showed how certain domains such as
friendships operated in both negative and positive ways and existed in dynamic interplay with related recovery domains such as
meaningful activities, for example, shopping together or going to the movies. Quotes and brief vignettes illustrated how these
domains commingled in an individual’s life and shifted over time (Padgett, Tran Smith, Choy-Brown, Tiderington, & Mercado,
2016).

Commentary: A simplified diagram of the study’s design would look like Figure 10.1. Thus, the study began as “big QUAL” in
approach (Step 1) but then evolved to accommodate a “small quant” study in which large amounts of qualitative data (over 4,000
pages of transcripts) were reduced to numerical ratings that were graphed (Step 2). Deep familiarity with the interviewees and the
data expedited this reductive process but also gave the ratings ecological validity. Step 3 involved examining specific recovery
domains for the 15 cases adjudicated as “changing” while going back and forth into the qualitative data. The purpose was to use both
sides to inform each other.

362
Examples of Mixed Methods Studies: Implementation and Action Research
Many types of research are amenable to mixed methods. To illustrate this point, the following paragraphs
feature examples drawn from two somewhat disparate areas: implementation science and action research. Two
examples from implementation research illustrate noncomplicated and complicated mixed methods designs.
With respect to the former, Couturier et al. (2015) conducted a multisite case study of four eating disorder
treatment centers that had received training in an innovative family-based treatment program. The researchers
tested the fidelity of the implementation through quantitative ratings of audio-recorded therapy sessions.
Concurrently, they conducted focus group interviews with clinicians and administrators to assess their
reactions to the new program. Results from both sources of data produced a more complete picture of how
family-based treatment was being received.

To illustrate a complicated design approach, Curry et al. (2015) describe a longitudinal concurrent mixed
methods implementation study to test whether changing hospital organizational culture might lead to better
outcomes for acute myocardial infarction (AMI) patients. The intervention, called Leadership Saves Lives
(LSL), was described in the journal Implementation Science in advance of being carried out in 10 hospitals.
Through tailored workshops and remote technical support with an array of hospital officials, LSL is intended
to introduce evidence-based practices that reduce response times and more efficiently target cardiac
procedures and medications used to treat heart attacks (AMIs).

According to Curry et al. (2015), qualitative and quantitative data are to be collected at each wave of data
collection over 24 months, the latter consisting of surveys assessing organizational culture as well as AMI
mortality rates. The study also relies heavily on key informant interviews with hospital leaders prior to each
LSL workshop and targeted field observation of patient rounds and cardiac team meetings. Addressing the
question of rigor, the authors state,

We will search for disconfirming evidence, interview multiple respondents at each hospital for
triangulation, and maintain a detailed audit trail to document analytic decisions. We will generate
thematic output which will describe recurrent and unifying concepts across the dataset, as well as
inform the generation of hypotheses for exploration in future studies. (Curry et al., 2015, p. 36)

Integration of the two sides would occur through triangulation and a search for disconfirming cases to assess
pre-post change over time. While the authors state that the study is “big QUAL” in emphasis, they
acknowledge that it has a deductive approach, that is, it entails testing hypotheses with a quantitative measure
of outcome (AMI mortality rates).

Action research and community-based participatory research (CBPR) have a natural affinity with mixed
methods in addressing goals of community improvement and social change. For example, a social justice
project was carried out in partnership with gay and bisexual men and transwomen in Bogota, Columbia to
understand processes of internal displacement due to violence, family conflict, and homonegativity (Zea,

363
Aguilar-Pardo, Betancourt, Reisen, & Gonzalez, 2014). Through sensitive but determined recruitment, the
researchers were able to enroll 42 participants for life history interviews using egalitarian dialogue. The
bilingual and binational (U.S. and Columbian) researchers jointly coded the interview data and used memos
to identify important themes in marginalization and stigma. In a sequential mixed methods design, the
qualitative findings were used to inform the development of a survey of attitudes and practices related to HIV
risk among gay, bisexual, and transgender (GBT) men. Survey respondents, sought from the gay male
community of Bogota, were recruited through respondent-driven sampling to achieve a survey sample of
1,000 (Zea et al., 2014).

In demonstrating interactions between the two sides, the authors note that the survey data showed a
statistically significant relationship between sex work and internal displacement. They juxtaposed this finding
with qualitative data showing that sex work was viewed as a positive experience for gay men seeking financial
independence and, for transgender men, an opportunity to enact feminine identity. Similarly, low rates of
HIV testing in the surveys were contextualized in the qualitative interviews as the result of fears of disclosure
and moderate concern about the dangers of HIV (Zea et al., 2014).

In an example of sequential mixed methods in CBPR, researchers in North Carolina partnered with 60
African American churches to launch the Partnership to Reach African Americans to Increase Smart Eating
(PRAISE) (Ammerman et al., 2003). Working closely with pastors and church leaders, the university-based
researchers sought to carry out a randomized trial of a dietary education program. Their partnership
consultations determined the content of the intervention as well as data collection and interpretation. For
example, since random assignment denying half of the churches the intervention was viewed as a negative, a
delayed intervention aspect was added to ensure control group churches’ access to the PRAISE program at the
study’s end.

The sequential design began with focus groups to decide on the parameters of the nutrition education
program as well as the content of a post-hoc survey to evaluate its effectiveness. After the PRAISE program
ended, in-depth interviews were conducted to obtain feedback on the intervention as well as the partners’
experiences in carrying out research. Community partners’ comments affirmed the importance of mutual
respect. One noted that church members needed to “know how much you care and not how much you know.”
Another commented, “If you want to do your research, then the relationship has to go above the rest”
(Ammerman et al., 2003, p. 1726).

364
Writing the Mixed Methods Report
The Journal of Mixed Methods Research (JMMR), which began publishing in 2007, features empirical reports,
but it is also a much-needed forum for explicating the “third paradigm” in research. Both research exemplars
and instructional discussions are needed given the newness of mixed methods and the understandable
confusion surrounding the variety of combinations and touch points possible. This, in turn, heightens the
need for transparency when writing up a mixed methods report. Aside from adopting the QUAL and QUAN
shorthand notations, the options for doing this are largely confined to straightforward descriptions of what
happened and when.

As with mixed methods design decisions, writing up the report revolves around a few pivot points: (1)
whether the sides were used concurrently or sequentially; (2) whether one side was embedded within the
other; and (3) whether data transformation took place or both sides remained intact throughout. Perhaps not
surprisingly, sequential designs are usually easier to write up, as both sides were presumably kept intact.
Concurrent and embedded designs place a greater burden on demonstrating how and when the integration
was carried out.

If the two sides were kept intact, the findings can be presented side-by-side, numerical and thematic. The
researcher interprets them in tandem and discusses the degree to which they converge or diverge
(triangulation may be invoked if appropriate). If the mixing occurred in the form of data transformation, the
findings are presented in that format (typically, transformed qualitative data are folded into the statistical
results). Perhaps not surprising given the motivation to mix methods in the first place, researchers tend to
favor keeping the sides intact as long as possible to give each side its due.

The challenge of mixed methods write-ups lies in giving the reader a complete account of what happened.
This usually means a step-by-step description as well as a rationale for each step. Few journals make
accommodations for the longer reports necessitated by mixed methods studies (the JMMR is an exception in
this regard). Diagrams or figures can capture the more complex mixed methods designs pictorially thereby
making the word description more parsimonious. A template longitudinal sequential design oscillating
between the quantitative and qualitative sides is shown in Figure 10.2. Recalling the double helix of
Mendlinger and Cwikel (2008), these undulations mark decision points for integration.

Figure 10.2 An Oscillating Mixed Methods Design

365
366
Challenges for Mixed Methods Studies
The discerning reader may have surmised by now that carrying out mixed methods studies poses a few
challenges. For example, most researchers are trained in one or the other method but not both. Coursework
and training in mixed methods is still in an early stage, and the overwhelming dominance of quantitative
methods in higher education means that shortages of qualitative expertise are all too common. Having dual
competencies through a team effort can overcome this drawback, but the lone investigator remains at a
disadvantage.

Mixed methods studies also require dual outlays of time and resources to ensure that both sides are given
sufficient attention (Stange, Miller, Crabtree, O’Connor, & Zyzanski, 1994). Another logistical challenge
accompanies oversight of the two sides when their rhythms and phases unfold in different ways. Qualitative
data analyses, for example, start early in data collection and may result in going back out into the field for
further sampling and data collection. Meanwhile, the quantitative side is proceeding in linear fashion, waiting
out the data collection before beginning statistical analyses. The qualitative side, even though working with a
smaller sample, takes considerable amounts of time and resources for transcription and data analysis not
matched on the quantitative side. Given these tensions, the temptation to use a dominant–less dominant
design is strong, and the quantitative side often comes out ahead in these situations.

Because mixed methods imply combining within the confines of a single study, questions can arise regarding
when and where one study ends and another begins. In sequential designs, too much time elapsed may lead to
the conclusion that two separate studies were conducted. In concurrent designs, the sides may have few or no
interactions and integration is minimal, thus giving the appearance of two unrelated studies.

There are also questions about the adequacy and provenance of the data. This is a threshold issue, since
nesting one side within the other risks the former losing viability when overshadowed and underresourced.
The threshold of viability tends to be more forgiving on the qualitative side (quantitative studies need
statistical power and larger sample sizes), but any dominance by one side requires paying attention to viability.

Collecting qualitative data under heavily quantitative auspices raises doubts about its authenticity and richness
(Morse, 2005). Moreover, qualitizing quantitative data is not the mirror image of quantitizing qualitative data
(this sentence is an admitted mouthful). With the former, quantitative data are aggregated and clustered, and
the resulting categories are based on decontextualized data that are hardly comparable to categories inductively
derived from in-depth interviews and/or extensive observation. Similarly, a disservice is often done to deep
and rich data when they are quantitized. Perhaps the best lesson to come from qualitizing and quantitizing is
that these data transformations should be done with great caution and transparency.

The conundrum posed by triangulation also affects many a mixed methods study. When results from both
sides are in accord, the researcher concludes (perhaps prematurely) that her findings are confirmed. As
discussed in Chapter 8, the meaning of triangulation has been expanded beyond corroboration to include
completeness. The problem arises over what to do when the qualitative and quantitative findings are neither

367
convergent nor complementary. Some researchers present the two sets of findings and ask the reader (and
future researchers) to resolve the differences. Others use the discrepancy as an opportunity to inquire further,
first to ensure that each of the sides is not flawed or biased in some way and then to examine and even use the
discrepancy as an opportunity to broaden or revise the study.

Box 10.5 gives an example of a mixed methods study in which serendipitous findings from the qualitative side
broadened the study and extended its impact to include new perspectives on barriers to breast cancer
screening.

368
Box 10.5 Serendipitous Findings and Mixed Methods: An Example
From the Harlem Mammogram Study
The Harlem Mammogram Study was a mixed method (QUAN+qual) study examining why African American women delay in
responding to an abnormal mammogram (Kerner et al., 2003). Interviews included measures of health locus of control, fear of
cancer, beliefs about racism, insurance status, and psychological distress. The less-dominant qualitative portion consisted of
questions about their mammogram experiences and discussions of aging, racism, body image, and female sexuality. Data analyses
were carried out separately by the quantitative and qualitative members of the team who had the requisite expertise. Both sides
shared an interest in understanding delay (or timeliness) and maintained close contact as the analyses proceeded. The quantitative
side used a multivariate logistic regression model to predict time to diagnostic resolution (within 3 months or longer). The
qualitative side used a modified grounded theory approach to identify codes and themes.

The qualitative data were intended to bring greater depth and understanding (a supplementary role), thus corresponding to
“completeness” as a goal of triangulation. As it happened, they also contributed two serendipitous findings that would not have
emerged otherwise. The first of these (mentioned in Chapter 6) was the “air theory” of cancer, which was subsequently noted in the
literature (Freeman, 2004). This folk belief, which holds that opening the body surgically exposes it to air that can cause dormant
cancer cells to grow and spread, was volunteered by several women during the qualitative interviews. Its relevance to our interest in
delay was obvious because a surgical biopsy posed just such a threat. The second unexpected qualitative finding was about the
physical and psychological toll of repeated diagnostic tests among women who had had multiple abnormal mammograms. A
significant minority of women in the study had to undergo repeated and painful needle biopsies and other procedures (Padgett,
Yedidia, Kerner, & Mandelblatt, 2001).

The qualitative results were significant in ways that our multivariate model did not (and probably could not) take into account. How,
after all, could we have anticipated or measured “air theory”? The results also turned out to have more grab than the statistical
analyses that proved to be disappointingly thin. Contrary to our hypothesized expectations, the effects of income, insurance coverage,
and systemic barriers were not found to be statistically significant. Among 30 predictor variables tested in the multivariate model,
only the degree of mammogram abnormality and whether the patient was given information were significant (Kerner et al., 2003).
Had we omitted the qualitative portion of the study, the underperforming quantitative model would have been the study’s only
finding for dissemination.

369
Mixed Methods Approaches to Program and Practice Evaluation
The shared pragmatism of program evaluation and mixed methods makes for a productive synergy.
Qualitative evaluators may grapple with the same epistemological issues as other qualitative researchers. Yet
theirs is a task that must remain true to the reality of a specific program, practice, or policy. They must also
balance using theories in general with “program theory” specifically (Rallis & Rossman, 2003).

In a fundamental sense, evaluation is “finding the value” (Ruckdeschel, Earnshaw, & Firrek, 1994; Scriven,
1967), making judgments according to societal priorities in light of the program’s mission. Evaluation
researchers must work under constraints of time, resources, and competing ideas regarding what is of “value.”
Even when vested interest groups—policymakers, administrators, staff, and clients—agree on a program’s
goals, the evaluator must work under close scrutiny and risk offending one or another of the stakeholders.

There are several ways that qualitative research can contribute to mixed methods evaluations (Greene, 2015;
Padgett, 2015; Rallis & Rossman, 2003). Quantitative evaluations are good at establishing what works, but
qualitative evaluations help to understand how a program succeeds or fails. Although admittedly more time
consuming, qualitative methods are less intrusive and less demanding than an experimental trial (Perreault,
Pawliuk, Veilleux, & Rousseau, 2006). Qualitative researchers can fade into the woodwork and respond
nimbly to the ebb and flow of organizational life. The addition of qualitative methods to a quantitative
evaluation adds flexibility and depth (Drake et al., 1993).

Qualitative research is particularly well suited to formative and process evaluation. Hong and colleagues (2005)
used ethnography to conduct a formative evaluation of an HIV prevention program with injection drug users.
Their findings regarding miscommunications and cultural relevance were used to inform and improve the
intervention. The inner workings of many programs—the dynamic interplay of the actors, their differing
perceptions of events, and the effects of culture and gender—are difficult to anticipate and measure. Indeed,
there is no substitute for what can be learned from extended ethnographic observation of a program as a work
in progress (Felton, 2003).

Qualitative approaches also mesh well with social advocacy values in evaluation—they empower less powerful
stakeholders (clients, lower level staff, etc.) by giving their voices greater prominence in “finding the value” of
the program. A mixed methods approach cannot guarantee a successful evaluation, but it is likely to enhance
the depth and relevance of the findings.

370
Summary and Concluding Thoughts
The popularity of mixed methods as a “third paradigm” is higher than ever before. Mixed methods studies
depend upon achieving a threshold of viability to ensure both sides have something to contribute. This is a tall
order since the two sides need to be integrated such that the sum is greater than the parts. Moreover, a
number of methodological and logistical challenges need to be met, not the least of which are the additional
outlays of expertise, time, and resources needed to do justice to both sides.

This chapter began with the archetypal possibilities based on sequencing and dominance. It also offered
several examples of the complicated, even messy, mixed methods designs that can take place. Determining
what to mix from the qualitative and quantitative sides, when to mix them, and how to make the linkages
takes careful advance planning. Notwithstanding these challenges, mixed methods open the door to
illuminating contrasts, whether done to corroborate, complement, or expand knowledge.

371
Exercises
1. Browse The Journal of Mixed Methods Research (or another of your choice), and locate a mixed methods
study of interest. Did the authors provide a diagram of the design using the usual “qual” and “quan”
notation? If not, how would you diagram the study’s design?
2. Choose a program or practice of interest to your class discussion group, and talk through how it could
be evaluated using mixed methods. Describe your design, and then specify how you would carry out the
proposed evaluation. What types of quantitative and qualitative methods, techniques, or analyses would
you use?
3. Think of a topic for which mixed methods would not be a good fit. Why is this so?

372
Additional Readings
Castro, F. G., Kellison, J. G., Boyd, S. J., & Kopak, A. (2010). A methodology for conducting integrative
mixed methods research and data analysis. Journal of Mixed Methods Research, 4(4), 342–360.

Cohen, D., & Crabtree, B. (2008). Robert Wood Johnson, Qualitative research guidelines project. Retrieved from
http://www.qualres.org

Creswell, J. W., Klassen, A. C., Plano Clark, V. L., & Smith, K. C. (2011, August). Best practices for mixed
methods research in the health sciences. Retrieved from https://obssr-
archive.od.nih.gov/mixed_methods_research/

Creswell, J. W., & Plano Clark, V. L. (2011). Designing and conducting mixed methods research. (2nd ed.).
Thousand Oaks, CA: Sage.

Curry, L. A., Nembhard, I. M., & Bradley, E. H. (2009). Qualitative and mixed methods provide unique
contributions to outcomes research. Circulation, 119, 1442–1452.

Greene, J. C. (2007). Mixed methods in social inquiry. San Francisco, CA: Jossey-Bass.

Guetterman, T., Creswell, J. T., & Kuckartz, U. (2015). Using joint displays and MAXQDA software to
present the results of mixed methods research. In M.T. McCrudden & G. Shraw (Eds.), Use of visual displays
in research and testing (pp. 145-175). Charlotte, NC: Information Age.

Johnson, P. J., & Onwuegbuzie, J. A. (2004). Mixed methods research: A research paradigm whose time has
come. Educational Researcher, 33(7), 14–26.

Miller, S. I., & Fredericks, M. (2006). Mixed-methods and evaluation research: Trends and issues.
Qualitative Health Research, 16, 567–579.

O’Cathain, A. (2010). Assessing the quality of mixed methods research: Toward a comprehensive framework.
In A. Tashakkori & C. Teddlie (Eds.), SAGE handbook on mixed methods research in the behavioral and social
sciences (2nd ed.; pp. 531–555). Thousand Oaks, CA: Sage.

Palinkas, L. A., Horwitz, S. M., Chamberlain, P., Hurlburt, M. S., & Landsverk, J. (2011). Mixed-methods
designs in mental health services research: A review. Psychiatric Services, 62(3), 255–263.

Plano Clark, V. L. (2010). The adoption and practice of mixed methods: U.S. trends in federally funded
health-related research. Qualitative Inquiry, 16(6), 428–440.

Stange, K. C., Crabtree, B. F., & Miller, W. L. (2006). Publishing multi-method research. Annals of Family
Medicine, 4, 292–294.

Tashakkori, A., & Creswell, J. W. (2007). The new era of mixed methods. Journal of Mixed Methods Research,
1(1), 3–7.

373
Tashakkori, A., & Teddlie, C. (Eds.). (2010). Handbook of mixed methods in social and behavioral research (2nd
ed.). Thousand Oaks, CA: Sage.

374
Appendix: Writing a Qualitative Methods Proposal for External
Funding

With their flexible and rather unpredictable nature, qualitative methods present special challenges when put
into proposal form. The bar is raised the highest when it comes to external funding (Ungar, 2006). Because
this book has been dedicated to conducting the type of rigorous qualitative research that can meet the
standards of external funders, I will focus in this appendix on the particulars of proposal writing and
submission. This discussion is largely based on my own experience in collaboratively obtaining grants from the
National Cancer Institute, the Centers for Disease Control (CDC), and the National Institute of Mental
Health (NIMH). I also draw on my experience in reviewing proposals and the experiences of successful
colleagues. I acknowledge at the outset a bias toward NIH funding because this is where most of my
experience lies. Foundations and other nongovernmental sources are also valuable supporters of high-quality
research.

The good news is most qualitative studies are low tech and low threshold and thus do not cost much if one
does not factor in time and labor. Meanwhile, research infrastructure supports within schools and departments
are on the rise, including workload reductions, methods consultants, small grants for training or pilot studies,
and staff assistance in preparing grant proposals. The annual conference of the Society for Social Work and
Research (SSWR) regularly offers workshops, panels, and sessions featuring qualitative methods. The
International Institute for Qualitative Methodology at the University of Alberta offers annual conferences
featuring workshops as well as lectures on qualitative methods (http://www.iiqm.ualberta.ca). A company run
by qualitative researchers (http://www.researchtalk.com) offers an array of workshops and summer intensive
courses taught by leading methodologists.

Obtaining external funding offers the luxury of having more resources such as personnel, participant
incentives, the latest software, and transcription services. It brings prestige to one’s institution and career
advancement for the investigator. The ability to succeed in the competitive arena of external research funding
has become a common expectation of faculty, graduate students, and program administrators. One can assume
that a proposal will receive close scrutiny, so thorough preparation begins months in advance. Writing a
research proposal is an extended effort in persuasion and confidence building in the researcher as well as the
methods.

375
Finding the Right Fit With Funding Priorities
All of the points made in Chapter 1 apply here regarding reasons for doing qualitative research, but first and
foremost the topic must be compelling and fill a gap in knowledge. It should also have wide and deep
relevance and foster innovative programs, treatments, and interventions. While private philanthropic
foundations have more leeway than government entities, they tend to favor known institutions over
individuals and are highly selective in keeping with their mission. The Bill and Melinda Gates Foundation,
for example, is generous in funding technology and global health but not social or behavioral programs. In my
research area of homelessness, the Melville Charitable Trust and the Oak Foundation have been in the
forefront. Similarly, there are many foundations that focus on child welfare and varied health problems. A
helpful clearinghouse on foundation support is the Foundation Center at http://foundationcenter.org.

Knowing the literature helps to shape the choice of research topic into something that a funder might want to
invest in. Although most research topics fall somewhere in the middle, some prompt an immediate “aha”
response as timely and urgent, and others appear to be nearly saturated with attention. Even among the latter,
it is frequently the case that an insider perspective is missing. Such a bottom-up approach opens the door to
qualitative methods and balances what is often an entirely top-down perspective on a social or human
problem.

It is useful to distinguish between research that is theory driven and knowledge generating and applied
research such as program evaluation. The actual difference between these two, which is often murky, can
hinge on how a study is framed and the funder’s overriding interest. NIH funded research ranges from DNA
sequencing to laboratory experiments to services research; a thread running through these is the need to be
theory driven, or scientific. One can go to the Research Portfolio Online Reporting Tools (RePORT) website
(http://projectreporter.nih.gov/reporter.cfm) and examine abstracts from successfully funded proposals to get
inspiration and information.

In contrast, federal entities such as the departments of Housing and Urban Development (HUD), SAMHSA
(Substance Abuse and Mental Health Services Administration), and the Department of Justice (DOJ) are
devoted to program development and evaluation. A quick glance at their websites reveals their interests in
assessing a variety of programs from jail diversion to suicide prevention. However, these government agencies
rarely open their research contract bidding to independent investigators (in direct contrast to NIH’s
extramural research portfolio).

Given the scope of federal agency interests and the many millions of dollars available for successful bidders,
large-budget research contractors (Westat, Inc., Abt Associates, Inc., Mathematica, Inc., Research Triangle
Institute, Rand Corporation) dominate this landscape. Humorously known as “Beltway Bandits” (the beltway
is the highway that surrounds Washington, DC), these consulting and research companies are able to quickly
draft research proposals to compete for the multimillion dollar projects that sustain them.

These days, qualitative studies are also likely to be funded as part of mixed methods projects. There are always

376
exceptions to prove this rule, for example, studies of sex workers and HIV, insider experiences of early onset
psychosis, or ways of recruitment into (and exits out of) street gangs. As discussed in Chapter 10, mixed
methods are not for everyone. However, when they are a good fit, they have the same synergizing appeal to
funders that they do more broadly.

377
Getting Past the Catch–22
Seasoned researchers usually do better at obtaining external funding than their less experienced counterparts
(although seniority is no guarantee of success). Two simple reasons account for this. First is the Catch–22 that
comes when review committees give higher ratings to proven researchers (on the assumption that experience
and qualifications will enhance the study’s probability of success). Second, mature researchers are usually
better able to put together a credible, strong proposal drawing on their knowledge of what it takes.

How does a novice overcome these drawbacks? New investigators can succeed by taking a few strategically
planned actions. First, they can seek the mentorship of senior researchers in their area who have successful
funding track records. (Ideally, one’s mentors are included in the proposal as co-investigators, consultants,
etc.) Second, they can seek out successful proposals in their areas of interest and emulate them. There are
many things that can be learned this way, including how the argument is made, how much emphasis to place
on methods, what a budget should look like, how to handle ethical issues, proper use of terminology, and so
forth.

While financial support for doctoral research has been growing, it usually consists of small grants to cover
direct costs such as transportation, incentives, and software. This appears to be a tacit carryover from the
sciences where one’s doctoral research is conducted in a professor’s lab (thus incurring little added expense). A
number of entities offer training and research opportunities to early career investigators including pre- and
postdoctoral fellowships. For example, the National Institutes of Health offer the “K” award to support early
career investigators through training and research funding.

378
Setting the Stage
In seeking external funding for a study, feasibility is critical. This may entail good working relationships with
gatekeepers whose help is pivotal for sample recruitment. Sampling needs should be met within a reasonable
time period and verified through support letters from the cooperating agency or program. Although large-
budget grants often include site fees to help offset the costs and disruption of a study, most site-specific
studies depend upon good will and transparency.

Pilot or preliminary study findings enhance confidence in the researcher’s commitment and the study’s
feasibility. The more this preparatory work can be incorporated into the proposal, the better. Finally, although
it may seem obvious, the researcher should corral all necessary proposal materials as early as possible. These
can include support letters, appendices, departmental approval letters, bio-sketches of key personnel, and so
forth. Waiting until the last minute will add enormous strain to the nail-biting process of assembling all of the
proposal’s parts and getting them in on time.

379
Questions From Proposal Reviewers
A qualitative methods proposal has a better chance than ever before, but it is still likely to raise questions
about methods (some more informed than others). Here, more than ever, thoroughness and specificity are
needed. The review of grant proposals varies somewhat depending on the funding source, but it invariably
involves evaluating the study’s overall significance and its methods. Typically, a panel of experienced
researchers review and rate the proposals, then meet to discuss their merits and recommend those deemed
worthy of funding. Thankfully, recent years have witnessed the inclusion of qualitative researchers on review
panels; the newest frontier is ensuring reviewers have sufficient expertise in mixed methods.

The following are some of the questions reviewers might raise during the review:

Do the principal investigator and the research team have the ability to make this study a success?
Is this study really filling a gap in knowledge or offering only marginal utility?
Is this study innovative?
Are there ethical problems that go unaddressed?
Is this study feasible given the time and money being requested?
Are the methods rigorous?
Is the sampling strategy biased?
Do the study findings have direct implications for practice/policy/future research?
How will the study contribute to theory development or wider knowledge about the phenomenon?

380
Budgeting Time and Resources
Forecasting the required amount of time and resources—and their associated costs—imposes a degree of
foreknowledge and precision atypical of qualitative inquiry. Yet formulas can be calculated based on desired
sample sizes and generally known information on the time it takes to conduct and transcribe interviews,
equipment and software costs, incentive payments, and so forth. Ultimately, research budgets must reflect the
study’s proposed reach and available resources.

Budget categories include personnel salaries and fringe benefits (investigators, interviewers, research assistants,
etc.), equipment (audio recorders, electronic notebook or tablets, desktop computers), research supplies
(software, batteries, flash memory drives, cloud storage, postage, etc.), transportation costs (subway/bus fares or
mileage for automobile travel), participant incentives and services (transcription, expert consultant fees), and
travel to conferences for presentations. Site fees may be paid to compensate for staff assistance (in recruitment
and data collection). If allowed, indirect costs (also called overhead) are computed as a percentage of the direct
costs to pay for office rent, clerical help, utilities, and so on. Allocations within and across these categories
must reflect a realistic appraisal of what is needed, that is, neither wildly inflated nor unrealistically under
budgeted.

Researchers should adhere to extant norms about incentive payments, consultant pay rates, and transcription
costs. When budgeting for digital audio recorders, go for a midrange price (the acuity of the recording will pay
for itself). Smartphone technology offers both audio and video recording although not as easily downloaded
for computer-based transcription. Audio and video files take up many megabytes of space, so cloud or external
drive storage is recommended.

Budgeting for interviews starts from the ground up. As a hypothetical example, assume that the study involves
two 90-minute interviews with 50 participants and transcription costs are $15 an hour. This amounts to 100
interviews times 1.5 hours or 150 hours of audio needing transcription. Because the ratio of recording hours to
transcription hours is anywhere from 1 to 3 to 1 to 8, it would be reasonable to allocate funds for 150
(recording) hours times 5 (transcription) hours, or 750 hours total for transcription time. At a rate of $15 per
hour, this equals $11,250 for transcription. Incentives at $25 per interview (50 participants, each having two
interviews) equal $2,500, and so forth.

These calculations should be fully detailed in the budget justification. As such, they vividly illustrate the
backstage labors involved in qualitative studies and offer a rationale for generously funding even a “small
sample” study. Indeed, the budget justification can help a proposal considerably by demonstrating how intense
and time-consuming rigorous qualitative studies can be.

381
Twenty Tips for Writing the Proposal
The following guidelines or tips can be turned into a handy checklist for planning and writing the proposal.

1. Start with compelling, unanswered study question(s) that will impress the reviewers.
2. Make sure the study begins with specific aims and research questions that present a clear road map for
the study
3. Write a concise but comprehensive abstract. (Remember that this may be the only description of your
study disseminated later on.)
4. Include a rationale for use of qualitative methods and their requisite strengths. (This is an unfair
disadvantage compared to quantitative proposals but is still recommended.)
5. Ground the study in a coherent, strong theoretical and conceptual framework.
6. Make sure that ethical guidelines are followed and potential threats to human subjects are minimized.
7. If the budget allows, include expert consultants who can provide guidance on methods or other
specialized study needs.
8. Be explicit about research design, including an explanation for why qualitative designs need to be
flexible and iterative.
9. Describe in detail the sampling and recruitment plans.
10. Build in a pilot study to ensure that protocols get a trial run through.
11. Describe procedures for retaining study participants (contact information, payment of incentives, etc.).
12. Structure the data analysis section around each specific aim or study goal and describe step by step how
the analyses will be carried out.
13. Include in each data analysis section a subsection titled “Strategies for Rigor” and indicate which will be
used (as appropriate).
14. Include a table in the proposal with some version of the following columns: Specific Aims (or Research
Questions), Sample Size and Source Sampling Criteria and Recruitment, Sources of Data, Type of Data
Analysis, Timeframe. Reading across, each row is a specific aim followed by its requisite information.
15. Include a timeline showing the length of the study (in months) on the x axis and each task stacked up
along the y axis. Lines and arrows show each task’s beginning and ending point.
16. If appropriate (and this is often the case), include a community advisory board (CAB) in the proposal. A
CAB provides valuable input and is an indication that the study is (and will be) responsive.
17. Make sure the budget is reasonable but not skimpy—feasibility and credibility matter.
18. Observe page limits, font sizes, and margins carefully. Ignoring these might result in a returned (or
discarded) proposal.
19. If appropriate, include plans for dissemination. Funders might want to be a part of this process to
varying degrees, but all have a vested interest in seeing the study’s conclusions influence practice, policy,
and future research.
20. Be patient and prepared to submit again (or somewhere else)!

382
383
Additional Readings
Creswell, J. W., Klassen, A. C., Plano Clark, V. L., & Smith, K. C. (2011, August). Best practices for mixed
methods research in the health sciences. Retrieved from
http://obssr.od.nih.gov/mixedmethods_research/pdf/Best_Practices_for_Mixed_Methods_Research.pdf

Locke L., Spirduso, W., & Silverman, S. (2013). Preparation of proposals for qualitative research: Different
assumptions. In L. Locke & W. Spirduso (Eds.), Proposals That Work: A Guide for Planning Dissertations and
Grant Proposals (6th ed., pp. 91–120). Thousand Oaks, CA: Sage.

Munhall, P., & Chenail, R. (2007). Qualitative research proposals and reports: A guide. Sudbury, MA: Jones &
Bartlett.

National Institutes of Health. (1999). Qualitative methods in health research: Opportunities and considerations in
application and review. Office of Behavioral and Social Science Research (OBSSR). Retrieved from
https://obssr-archive.od.nih.gov/pdf/Qualitative.PDF

Ungar, M. (2006). “Too ambitious”: What happens when funders under-estimate the strength of qualitative
research design. Qualitative Social Work, 5(2), 261–277.

384
References

Abrams, L. S., & Anderson-Nathe, B. (2013). Compassionate confinement: A year in the life of Unit C. New
Brunswick, NJ: Rutgers University.

Adams, V. (2013). Evidence-based global public health: Subjects, profits, erasures. In J. Biehl & A. Petryna
(Eds.), When people come first (pp. 54–90). Princeton, NJ: Princeton University Press.

Agar, M. H. (1980). The professional stranger: An informal introduction to ethnography. New York, NY:
Academic Press.

Agar, M. (2006). An ethnography by any other name. Forum Qualitative Sozialforschung/Forum: Qualitative
Social Research, 7(4), Art. 36, http://nbn-resolving.de/urn:nbn:de:0114-fqs0604367

Agar, M. H., & McDonald, J. (1995). Focus groups and ethnography. Human Organization, 54(1), 78–86.

Alderson, P., & Morrow, V. (2004). Ethics, social research and consulting with children and young people.
Ilford: Barnardo’s.

Alexander, M. (2010). The new Jim Crow: Mass incarceration in the age of color-blindness. New York, NY:
New Press.

Alexander, M., & West, C. (2012). The new Jim Crow: Mass incarceration in an age of color-blindness. New
York, NY: The New Press.

Allison, K. R., & Rootman, I. (1996). Scientific rigor and community participation in health promotion
research: Are they compatible? Health Promotion International, 11(4), 333–340.

Altheide, D. L., & Johnson, J. M. (1994). Criteria for assessing interpretive validity in qualitative research. In
N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 485–499). Thousand Oaks,
CA: Sage.

Alvesson, M. (2011). Interpreting interviews. London, UK: Sage.

Alvesson, M., & Karreman, D. (2011). Qualitative research and theory development. London, UK: Sage.

385
Ammerman, A., Corbie-Smith, G., St. George, D. J., Washington, C., Weathers, B., & Jackson-Christian,
B. (2003). Research expectations among African American church leaders in the PRAISE! Project: A
randomized trial guided by community-based participatory research. American Journal of Public Health,
93, 1720–1727.

Anderson, E. (2000). Code of the street. New York, NY: W.W. Norton.

Andrews, M. (2014). Narrative imagination and everyday life. Oxford, UK: Oxford University Press.

Anfara, V. A., & Mertz, N. T. (Eds.). (2006). Theoretical frameworks in qualitative research. Thousand
Oaks, CA: Sage.

Angell, B., & Bolden, G. B. (online, April 29, 2015). Justifying medication decisions in mental health care:
Psychiatrists’ accounts for treatment recommendations. Social Science & Medicine.

Annells, M. (1996). Grounded theory method: Philosophical perspectives, paradigm of research, and
postmodernism. Qualitative Health Research, 6(3), 379–393.

Annells, M. (2006). Triangulation of qualitative approaches: Phenomenology and grounded theory. Journal of
Advanced Nursing, 56(1), 55–61.

Armour, M., Rivaux, S. L., & Bell, H. (2009). Using context to build rigor: Application of two hermeneutic
phenomenological studies. Qualitative Social Work, 8, 101–122.

Atkinson, P. (2005, September). Qualitative research—Unity and diversity [25 paragraphs]. Forum
Qualitative Sozialforschung/Forum: Qualitative Social Research, 6(3), Article 26. Retrieved from
http://www.qualitative-research.net/fqs-texte/3–05/05–3-26-e.htm

Atkinson, P., & Coffey, A. (2002). Revisiting the relationship between participant observation and
interviewing. In J. Gubrium & J. A. Holstein (Eds.), Handbook of interview research (pp. 88–110).
Thousand Oaks, CA: Sage.

Atkinson, P., & Silverman, D. (1997). Kundera’s immortality: The interview society and the invention of the
self. Qualitative Inquiry, 3, 304–325.

Ayling, R., & Mewse, A. J. (2009). Evaluating Internet interviews with gay men. Qualitative Health

386
 Research, 19, 566–576.

Baker, C., Wuest, J., & Stern, P. N. (1992). Method slurring: The grounded theory/phenomenology example.
Journal of Advanced Nursing, 17, 1335–1360.

Baker, S. E., & Edwards, R. (2012). How many qualitative interviews is enough? London, UK: National
Centre for Research Methods. Retrieved from http://eprints.ncrm.ac.uk/2273/4/how_many_interviews.pdf

Bamberg, M. (2004). Considering counter narratives. In M. Bamberg & M. Andrews (Eds.), Considering
counter-narratives: Narrating, resisting, making sense (pp. 351–371). Amsterdam: John Benjamins.

Barbour, R. S., & Barbour, M. (2003). Evaluating and synthesizing qualitative research: The need to develop
a distinctive approach. Journal of Evaluation in Clinical Practice, 9, 179–186.

Barker, J., & Weller, S. (2003). ‘Is it fun?’ Developing children-centered research methods. International
Journal of Sociology and Social Policy, 23(1), 33–58.

Barkin, S., Ryan, G., & Gelberg, L. (1999). What pediatricians can do to further violence prevention: A
qualitative study. Injury Prevention, 5, 53–58.

Barnett-Page, E., & Thomas, J. (2009). Methods for the synthesis of qualitative research: A critical review.
ESRC National Centre for Research Methods. Working Paper Series 01–09. Retrieved from
http://eppi.ioe.ac.uk/cms/Default.aspx?tabid=188

Bazeley, P. (2010). Review of the mixed methods reader. Journal of Mixed Methods Research, 4(1), 79–81.

Beck, C. T. (1993). Teetering on the edge: A substantive theory of postpartum depression. Nursing Research,
42(1), 42–50.

Becker, H. (1996). The epistemology of qualitative research. In R. Jessor, A. Colby, & R. Schweder (Eds.),
Ethnography and human development (pp. 53–72). Chicago, IL: University of Chicago Press.

Becker, H. (1998). Tricks of the trade: How to think about your research while you’re doing it. Chicago, IL:
University of Chicago Press.

387
Becker, H., Geer, B., Hughes, E., & Strauss, A. (1961). Boys in white: Student culture in medical school.
Chicago, IL: University of Chicago Press.

Beebe, J. (2002). Rapid assessment process. Landham, MD: Altamira Press.

Berelson, B. (1952). Content analysis in communication research. Glencoe, IL: Free Press.

Berger, P., & Luckmann, T. (1967). The social construction of reality. New York, NY: Doubleday.

Berguno, G., Leroux, P., McAinsh, K., & Shaikh, S. (2004). Children’s experience of loneliness at school and
its relation to bullying and the quality of teacher interventions. The Qualitative Report. 9, 483–499.

Bernard, H. R., & Gravelee, C. C. (Ed.). (2014). Handbook of methods in cultural anthropology (2nd ed.).
Walnut Creek, CA: Altamira Press.

Bernard, H. R., & Ryan, G. (2010). Analyzing qualitative data: Systematic approaches. Thousand Oaks, CA:
Sage.

Biehl, J. (2013). Vita: Life in a zone of social abandonment. Berkeley, CA: University of California Press.

Blix, S. B., & Wettergren, A. (2015). The emotional labor of gaining and maintaining access to the field.
Qualitative Research, 15, 688–704.

Bloor, M. (2001). Techniques of validation in qualitative research: A critical commentary. In R. M. Emerson

(Ed.), Contemporary field research (pp. 383–396). Prospect Heights, IL: Waveland Press.

Blumer, A. (1969). Symbolic interactionism: Perspective and method. Englewood Cliffs, NJ: Prentice-Hall.

Boeri, M. W. (2004). “Hell I’m an addict, but I ain’t no junkie”: An ethnographic analysis of aging heroin
users. Human Organization, 63(2), 236–246.

Bohm , A. (2004). Theoretical coding: Text analysis in grounded theory. In U. Flick, E. von Kardorff, & I.
Steinke (Eds.), A companion to qualitative research (pp. 270–275). London, UK: Sage.

Borgatti, S. P. (1994). Cultural domain analysis. Journal of Quantitative Anthropology, 4, 261–278.

388
Bourdieu, P. (1999). Understanding. In P. Bourdieu (Ed.), The weight of the world: Social suffering in
contemporary society (pp. 607–626). Cambridge, MA: Polity Press.

Bowen, G. A. (2006). Grounded theory and sensitizing concepts. International Journal of Qualitative
Methods, 5(3), Article 2.

Bowen, G. (2008). Naturalistic inquiry and the saturation concept: A research note. Qualitative Research, 8,
137–152.

Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9,
27–40.

Bowen, G. A. (2010). From qualitative dissertation to quality articles: Seven lessons learned. The Qualitative
Report, 15(4), 864–879. Retrieved from http://nsuworks.nova.edu/tqr/vol15/iss4/6

Boyatzis, R. E. (1998). Transforming qualitative information: Thematic analysis and code development.
Thousand Oaks, CA: Sage.

Bradshaw, T. K. (1999). Communities not fazed: Why military base closures may not be catastrophic. Journal
of American Planning Association, 65, 193–206.

Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3,
77–101

Brekke, J. (2012). Shaping a science of social work. Research on Social Work Practice, 22, 455–464.

Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press.

Bronfenbrenner, U. (2004). Making human beings human: Bio-ecological perspectives on human

development. Thousand Oaks, CA: Sage.

Bryman, A. (2006). Integrating quantitative and qualitative research: How is it done? Qualitative Research,
6(1), 97–113.

Burawoy, Michael. (1998). The extended case method. Sociological Theory 16, 4–34.

389
Butler, A. E., Hall, H., Willetts, G., & Copnell, B. (2015). Family experience and PICU death: A meta-
synthesis. Pediatrics, 136, 298–313.

Cabassa L. J., Parcesepe, A., Nicasio, A., Baxter, E., Tsemberis, S., & Lewis-Fernández. R. (2013). Health
and wellness photovoice project: Engaging consumers with serious mental illness in health care
interventions. Qualitative Health Research, 23, 618–630.

Cameron, W. B. (1963). Informal sociology. New York, NY: Random House.

Campbell, M. L. (2004). Mapping social relations: An introduction to institutional ethnography. Walnut

Creek, CA: Altamira Press

Campbell, R., & Arens, C. E. (1998). Innovative community services for rape victims: An application of
multiple case study methodology. American Journal of Community Psychology, 26(4), 537–571.

Carmack, H. (2014). A cycle of redemption in a medical error disclosure and apology program. Qualitative
Health Research, 24, 860–869.

Charmaz, K. (2006). Constructing grounded theory. Thousand Oaks, CA: Sage.

Charmaz, K. (2014). Constructing grounded theory (2nd ed.). Thousand Oaks, CA: Sage.

Cherryholmes, C. H. (1992). Notes on pragmatism and scientific realism. Educational Researcher, 14, 13–17.

Choy-Brown, M., Padgett, D., Tran Smith, B., & Tiderington, D. (2015). Sorting it out: Eliciting consumer
priorities on recovery in supported housing. American Journal of Psychiatric Rehabilitation.

Clandinin, D. J., & Connelly, F. M. (2004). Narrative inquiry: Experience and story in qualitative research.
San Franscisco, CA: Jossey Bass.

Clark, J. P. (2003). How to peer review a qualitative manuscript. In F. Godlee & T. Jefferson (Eds.). Peer
review in health sciences. (2nd ed.; pp. 219–235). London, UK: BMJ Books.

Clarke, A. (2005). Situational analysis: Grounded theory after the postmodern turn. Thousand Oaks, CA:
Sage.

390
Clifford, J., & Marcus, G. E. (Eds.). (1986). Writing culture: The poetics and politics of ethnography.
Berkeley: University of California Press.

Clough, P. T. (1998). The end(s) of ethnography: From realism to social criticism (2nd ed.). New York, NY:
Peter Lang.

Coffey, A., & Atkinson, P. (1996). Making sense of qualitative data. Thousand Oaks, CA: Sage.

Cohen, D., & Crabtree, B. J. (2008). Evaluative criteria for qualitative research in health care: Controversies
and recommendations. Annals of Family Medicine, 6, 331–339.

Cohen, M. Z., Kahn, D. L., & Steeves, R. H. (2000). Hermeneutic phenomenological research: A practical
guide for nurse researchers. Thousand Oaks, CA: Sage.

Colaizzi, P. F. (1978). Psychological research as the phenomenologist views it. In R. Valle & M. King (Eds.),
Existential-phenomenological alternatives for psychology (pp. 48–71). New York, NY: Oxford University
Press.

Collier, J., & Collier, M. (1986). Visual anthropology: Photography as a research method. Albuquerque:
University of New Mexico Press.

Connolly, A. (2008). Challenges of generating qualitative data with socially excluded young people.
International Journal of Social Research Methodology, 22, 201–214.

Cook, T. D., & Reichardt, C. S. (Eds.). (1979). Qualitative and quantitative methods in evaluation research.
Beverly Hills, CA: Sage.

Corbin, J., & Strauss, A. (2007). Basics of qualitative research (3rd ed.). Thousand Oaks, CA: Sage.

Cornwall, A., & Jewkes, R. (1995). What is participatory research? Social Science & Medicine, 41(12),
1667–1676.

Couturier, J., Kimber, M., Lock, J., Barwick, M., McVey, G., Findlay, S. . . . Woodford, T. (2015).
Implementing highly specialized and evidence-based pediatric eating disorder treatment: Protocol for a
mixed methods evaluation. Implementation Science, 10, 40–47.

391
Crabtree, B. F., & Miller, W. L. (1999). Doing qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Cressey, D. E. (1953). Other people’s money: A study in the social psychology of embezzlement. Glencoe,
IL: Free Press.

Creswell, J. W. (2013). Qualitative inquiry and research design (3rd ed.). Thousand Oaks, CA: Sage.

Creswell, J. W., Klassen, A. C., Plano Clark, V. L., & Smith, K. C. (2011, August). Best practices for mixed
methods research in the health sciences. Retrieved from
http://obssr.od.nih.gov/mixed_methods_research/pdf/Best_Practices_for_Mixed_Methods_Research.pdf

Creswell, J. W., & Plano Clark, V. L. (2011). Designing and conducting mixed methods research. (2nd ed.).
Thousand Oaks, CA: Sage.

Curry, L. A., Linnander, E. L., Brewster, A. L., Ting, H., Krumholz, H., & Bradley, E. H. (2015).
Organizational culture change in U.S. hospitals: A mixed methods longitudinal intervention study.
Implementation Science, 10, 27–39.

Czarniawska, B. (2004). Narratives in social science research. Thousand Oaks, CA: Sage.

Damschroder, L. J., Aron, D. C., Keith, R. E., Kirsh, S. R., Alexander, J. A., & Lowery, J. C. (2009).
Fostering implementation of health service research findings into practice. Implementation Science, 4,
50–65.

Danso, R. (2015). An integrated framework of critical cultural competence and anti-oppressive practice for
social justice social work research. Qualitative Social Work, 14, 572–588.

Delgado, R., & Stefancic, J. (2011). Critical race theory: An introduction. New York: NYU Press.

Denzin, N. K. (1978). The research act: A theoretical introduction to sociological methods (2nd ed.). New
York, NY: McGraw-Hill.

Denzin, N. K. (1989). Interpretive interactionism. Newbury Park, CA: Sage.

Denzin, N. K. (2003). Performance ethnography: Critical pedagogy and the politics of culture. Thousand

392
 Oaks, CA: Sage.

Denzin, N. K. (2008). The new paradigm dialogs and qualitative inquiry. International Journal of Qualitative
Studies in Education, 21, 315–325.

Denzin, N. K., & Giardina, M. D. (Eds.). (2006). Qualitative inquiry and the conservative challenge. Walnut
Creek, CA: Left Coast Press.

Denzin, N. K., & Lincoln, Y. S. (Eds.). (1994). Handbook of qualitative research. Thousand Oaks, CA:
Sage.

Denzin, N. K., & Lincoln, Y. S. (Eds.). (2011). Handbook of qualitative research (4th ed.). Thousand Oaks,
CA: Sage.

DeSilva, M. J., Harpham, T., Tuan, T., Bartolini, R., Penny, M. E., & Huttly, S. R. (2006). Psychometric
and cognitive validation of a social capital measurement tool in Peru and Vietnam. Social Science &
Medicine, 62, 941–953.

Dickson-Swift, V., James, E. L., Kippen, S., & Liamputtong, P. (2009). Researching sensitive topics:
Qualitative research as emotion work. Qualitative Research, 9, 61–79.

Diez-Roux, A. (1998). Bringing context back into epidemiology: Variables and fallacies in multi-level
analysis. American Journal of Public Health, 88, 216–222.

Dixon-Woods, M., Booth, A., & Sutton, A. J. (2007). Synthesizing qualitative research: A review of
published reports. Qualitative Research, 7(3), 375–422.

Donmoyer, R. (1990). Generalizability and the single-case study. In E. W. Eisner & A. Peshkin (Eds.),
Qualitative inquiry in education: The continuing debate (pp. 175–200). New York, NY: Teachers College
Press.

Drake, R. E., Bebout, R. R., Quimby, E., Teague, G. B., Harris, M., & Roach, J. P. (1993). Process
evaluation in the Washington D.C. Dual Diagnosis Project. Alcoholism Treatment Quarterly, 10,
113–124.

Drisko, J. (1997). Strengthening qualitative studies and reports: Standards to enhance academic integrity.

393
 Journal of Social Work Education, 33, 187–197.

Duneier, M. (1999). Sidewalk. New York, NY: Farrar, Straus, & Giroux.

Dupuis, A., & Thorns, D. C. (1998). Home, home ownership, and the search for ontological security. The
Sociological Review, 46(1), 24–47.

Eakin J. M., & Mykhalovskiy, E. (2003). Reframing the evaluation of qualitative health research: Reflections
on a review of appraisal guidelines in the health sciences. Journal of Evaluation in Clinical Practice, 9,
187–194.

Elder, G. (1994). Time, human agency and social change: Perspectives on the life course. Social Psychology
Quarterly, 57, 4–15.

Ellis, C., & Bochner, A. P. (2000). Autoethnography, personal narrative, reflexivity: Researcher as subject. In
N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 733–768). Thousand
Oaks, CA: Sage.

Elo, S., & Kyngäs, H. (2008). The qualitative content analysis process. Journal of Advanced Nursing. 62,
107–15.

Ely, M., Anzul, M., Friedman, T., Garner, D., & Steinmetz, A. M. (1991). Doing qualitative research:
Circles within circles. London, UK: Falmer.

Emerson, R. M., Fretz, R. I., & Shaw, L. L. (2011). Writing ethnographic fieldnotes, (2nd ed.). Chicago, IL:
University of Chicago Press.

Eng, E., Moore, K. S., Rhodes, S. D., Griffith, D. M., Allison, L. L., Shirah, K., & Mebane, E. M. (2005).
Insiders and outsiders assess who is “the community.” In Israel, B. A., Eng, E., Schulz, A. U., & Parker, E.
A. (Eds.). (2005). Methods in community-based participatory research for health (pp. 77–100). San
Francisco, CA: Jossey-Bass.

Erickson, F. (1986). Qualitative methods in research on teaching. In M. C. Wittrock (Ed.), Handbook of

research on teaching (3rd ed., pp. 119–161). New York, NY: Macmillan.

Erwin, E. J., Brotherson, M. J., & Summers, J. A. (2011). Understanding qualitative meta-synthesis: Issues

394
 and opportunities in early childhood intervention research. Journal of Early Intervention, 33, 186–200.

Esposito, N. (2001). From meaning to meaning: The influence of translation techniques on non-English
focus group research. Qualitative Health Research, 11(4), 568–579.

Estroff, S. (1981). Making it crazy. Berkeley: University of California Press.

Etherington, K. (2007). Working with traumatic stories: From transcriber to witness. International Journal of
Social Research Methodology 10, 85–97.

Fairclough, N. (2003). Analyzing discourse: Textual analysis for social research. New York, NY: Routledge.

Fals-Borda, O. (Ed.). (1998). People’s participation: Challenges ahead. New York, NY: Apex.

Fals-Borda, O., & Rahman, M. A. (Eds.). (1991). Action and knowledge: Breaking the monopoly with
participatory action research. New York, NY: Intermediate Technology/Apex.

Farmer, P. (1999). Infections and inequalities: The modern plagues. Berkeley, CA: University of California
Press.

Farnell, B., & Graham, L. R. (2000). Discourse-centered methods. In H. R. Bernard (Ed.), Handbook of
methods in cultural anthropology (pp. 411–454). Walnut Creek, CA: Altamira Press.

Feagin, J. R., Orum, A. M., & Sjoberg, G. (Eds.). (1991). A case for the case study. Chapel Hill: University
of North Carolina Press.

Felton, B. J. (2003). Defining location in the mental health system: A case study of a consumer-run agency.
American Journal of Community Psychology, 36, 373–386.

Fereday, J., & Muir-Cochrane, E. (2006). Demonstrating rigor using thematic analysis: A hybrid approach of
inductive and deductive coding and theme development. International Journal of Qualitative Methods,
5(1), Article 1.

Fernandez, J., & Huber, M. T. (2001). Irony in action: Anthropology, practice and the moral imagination.
Chicago, IL: University of Chicago Press.

395
Festinger, L., Reicken, H. W., & Schacter, S. (1956). When prophecy fails. New York, NY: Start.

Fetterman, D. M. (1989). Ethnography step by step. Newbury Park, CA: Sage.

Fetters, M. D., & Freshwater, D. (2015). The 1+1=3 integration challenge. Journal of Mixed Methods
Research, 9, 115–117.

Fine, G. A., & Martin, D. D. (1990). A partisan view: Sarcasm, satire, and irony as voices in Erving
Goffman’s Asylums. Journal of Contemporary Ethnography, 19, 89–115.

Flaherty, M. G. (2002). The “crisis” in representation: Reflections and assessments. Journal of Contemporary
Ethnography, 31(4), 508–516.

Flick, U. (2004). Triangulation in qualitative research. In U. Flick, E. von Kardorff, & I. Steinke (Eds.), A
companion to qualitative research (pp. 178–183). London, UK: Sage.

Flicker, S., Haans, D., & Skinner, H. (2004). Ethical dilemmas in research on Internet communities.
Qualitative Health Research, 14, 124–130.

Fonow, M. M., & Cook, J. A. (Eds.). (1991). Beyond methodology: Feminist scholarship as lived research.
Bloomington: Indiana University Press.

Fontana, A., & Frey, J. H. (1994). Interviewing: The art of science. In N. K. Denzin & Y. S. Lincoln (Eds.),
Handbook of qualitative research (pp. 361–376). Thousand Oaks, CA: Sage.

Fost, N., & Levine, R.J. (2007). The dysregulation of human subjects research. JAMA, 298, 2196–2198.

Foster-Fishman, P., Berkowitz, S. L., Lounsbury, D. W., Jacobson, S., & Allen, N. (2001). Building
collaborative capacity in community coalitions: A review and integrative framework. American Journal of
Community Psychology, 29(2), 241–261.

Foucault, M. (1976). The history of sexuality. London, UK: Penguin.

Foucault, M. (1980). Power/knowledge: Selected interviews and other writings, 1972–1977. New York, NY:
Pantheon.

396
Frank, A. W. (2010). Letting stories breathe. Chicago, IL: University of Chicago Press.

Frank, A. W. (2012). Practicing dialogical narrative analysis. In J. A. Holstein & J. F. Gubrium (eds.).
Varieties of narrative analysis (pp. 33–52). Thousand Oaks, CA: Sage.

Fraser, M. W. (1995). Rich, relevant, and rigorous: Do qualitative methods measure up? Social Work
Research, 19(1), 25–27.

Freeman, H. P. (2004). Poverty, culture and social injustice: Determinants of cancer disparities. Cancer: A
Cancer Journal for Clinicians, 54, 72–77.

Freire, P. (1973). Pedagogy of the oppressed. New York, NY: Seabury Press.

Freundlich, M., Avery, R. J., & Padgett, D. K. (2007). Care or scare: The safety of youth in congregate care
in New York City. Child Abuse and Neglect, 31(2), 173–186.

Gage, N. L. (1989). The paradigm wars and their aftermath: A historical sketch of research on teaching.
Educational Researcher, 18, 4–10.

Gair, S. (2002). In the thick of it: A reflective tale from an Australian social worker/qualitative researcher.
Qualitative Health Research, 12(1), 130–139.

Gamson, J. (2000), Sexualities, queer theory, and qualitative research. In N. Denzin & Y. Lincoln, Handbook
of Qualitative Research (2nd ed.; pp. 347–356). Thousand Oaks, CA: Sage.

Gee, J. P. (2005). An introduction to discourse analysis: Theory and method. London, UK: Routledge.

Geertz, C. (1973). The interpretation of cultures: Selected essays. New York, NY: Basic Books.

Geertz, C. (1988). Works and lives: The anthropologist as author. Stanford, CA: Stanford University Press.

Gergen, K. J. (2014). Pursuing excellence in qualitative inquiry. Qualitative Psychology, 1, 49–60.

Gergen, K. J., Josselson, R., & Freeman, M. (2015). The promises of qualitative inquiry. American
Psychologist, 70, 1–10.

397
Gerring, J. (2007). Case study research: Principles and practices. London, UK: Cambridge University Press.

Gibbs, L., & Gambrill, E. (2002). Evidence-based practice: Counter-arguments to objections. Research on
Social Work Practice, 12(3), 452–476.

Giddens, A. (1990). Consequences of modernity. Oxford, UK: Polity Press.

Giddings, L. S. (2006). Mixed-methods research: Positivism dressed in drag? Journal of Research in Nursing,
11, 195–203.

Gilgun, J. F. (1994). Hand into glove: Grounded theory and social work practice research. In W. Reid & E.
Sherman (Eds.), Qualitative methods and social work practice research pp. 115–125. New York, NY:
Columbia University Press.

Gilgun, J. F. (1995). We shared something special: The moral discourse of incest perpetrators. Journal of
Marriage and the Family, 57, 265–281.

Gilgun, J. F., & Abrams, L. (2002). The nature and usefulness of qualitative social work research. Qualitative
Social Work, 1, 39–55.

Gill, R., Barbour, J., & Dean, M. (2014). Shadowing in/as work: Ten recommendations for shadowing
fieldwork practice. Qualitative Research in Organizations and Management, 9, 69–99.

Giorgi, A. (Ed.). (1985). Phenomenology and psychological research. Pittsburgh, PA: Duquesne University
Press.

Giorgi, A. (2009). The descriptive phenomenological method in psychology. Pittsburgh, PA: Duquesne
University Press.

Glaser, B. (1978). Theoretical sensitivity. Mill Valley, CA: The Sociology Press.

Glaser, B. G. (2002). Conceptualization: On theory and theorizing using grounded theory. International
Journal of Qualitative Methods, 1(2), Article 3.

Glaser, B. G., & Strauss, A. L. (1965). Awareness of dying. New York, NY: Aldine.

398
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research.
Chicago, IL: Aldine.

Goetz, J., & LeCompte, M. (1984). Ethnography and qualitative design in educational research. Orlando,
FL: Academic Press.

Goffman, A. (2014). On the run: Fugitive life in an American city. Chicago, IL: University of Chicago Press.

Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Basic Books.

Goffman, E. (1961). Asylums: Essays on the social situation of mental patients and other inmates. Garden
City, NY: Basic Books.

Gooden, R. J., & Winefield, H. R. (2007). Breast and prostate cancer online discussion boards: A thematic
analysis of gender differences and similarities. Journal of Health Psychology, 12(1), 103–114.

Greene, J. C. (2000). Understanding social programs through evaluation. In N. K. Denzin & Y. S. Lincoln
(Eds.), Handbook of qualitative research (2nd ed., pp. 981–1000). Thousand Oaks, CA: Sage.

Greene, J. C. (2015). The emergence of mixing methods in the field of evaluation. Qualitative Health
Research, 25, 746–750.

Greene, J. C., Caracelli, V. J., & Graham, W. F. (1989). Toward a conceptual framework for mixed-method
evaluation designs. Educational Evaluation and Policy Analysis, 11(2), 255–274.

Greenhaugh T. (1997). How to read a paper: The basics of evidence based medicine. London, UK: BMJ.

Groenewald, T. (2004). A phenomenological research design illustrated. International Journal of Qualitative

Methods, 3(1), Article 4.

Grypdonck, M. (2006). Qualitative health research in the era of evidence-based practice. Qualitative Health
Research, 16, 1371–1385.

Guba, E. G., & Lincoln, Y. S. (1981). Effective evaluation. San Francisco, CA: Jossey-Bass.

399
Guba, E. G., & Lincoln, Y. S. (1989). Fourth generation evaluation. Newbury Park, CA: Sage.

Gubrium, J. F., & Holstein, J. A. (2000). Analyzing interpretive practice. In N. K. Denzin & Y. S. Lincoln
(Eds.), Handbook of qualitative research (pp. 487–508). Thousand Oaks, CA: Sage.

Guest, G. (2012). Describing mixed methods research: An alternative to typologies. Journal of Mixed
Methods Research, 7, 141–151.

Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data
saturation and variability. Field Methods, 18(1), 59–82.

Guest, G., & MacQueen, K. M. (2008). Handbook for team-based qualitative research. Walnut Creek, CA:
Altamira.

Hamilton, R. J., & Bowers, B. J. (2006). Internet recruitment and e-mail interviews in qualitative studies.
Qualitative Health Research, 16, 821–826.

Hammersley, M. (2004). Teaching qualitative method: Craft, profession or bricolage? In C. Seale, G. Gobo,
J. F. Gubrium, & D. Silverman (Eds.), Qualitative research practice (pp. 549–560). London, UK: Sage.

Hammersley, M. (2008). Questioning qualitative inquiry. London, UK: Sage.

Hammersley, M., & Gomm, R. (2008). Assessing the radical critique of interviews. In M. Hammersley,
Questioning qualitative inquiry (pp. 89–100). London, UK: Sage.

Hancock, D. R., & Algozzine, B. (2006). Doing case study research: A practical guide for beginning
researchers. New York, NY: Teachers College Press.

Haraway, D. (1988). Situated knowledge: The science question in feminism and the privilege of partial
perspectives, Feminist Studies, 14, 575–599.

Harding, S. (1987). Feminism and methodology. Bloomington, IN: Indiana University Press.

Harper, D. (2012). Visual sociology. London, UK: Routledge.

400
Hawkins, R. L., & Abrams, C. (2007). Disappearing acts: The social networks of formerly homeless
individuals with co-occurring disorders. Social Science & Medicine, 65, 2031–2042.

Heckathorn, D. D. (1997). Respondent-driven sampling: A new approach to study hidden populations. Social
Problems, 44, 174–199.

Henwood, B., & Padgett, D. K. (2007). The self-medication hypothesis revisited. American Journal on
Addictions, 16(3), 160–165.

Henwood, B. F., Padgett, D. K., Tran Smith, B., & Tiderington, E. (2012). Substance abuse recovery after
experiencing homelessness and mental illness: Case studies of change over time. Journal of Dual Diagnosis,
8, 238–246.

Hesse-Biber, S. N. (2004). Analysis, interpretation and the writing of qualitative data. In S. N. Hesse-Biber
& P. Leavy, Approaches to qualitative research: A reader on theory and practice (pp. 409–425), New York,
NY: Oxford.

Hesse-Biber, S. N. (2006). Handbook of feminist research: Theory and praxis. Thousand Oaks, CA: Sage.

Hesse-Biber, S. N. (2015). Mixed methods research: The “thingness” problem. Qualitative Health Research,
25, 775–788.

Hessler, R. M., Downing, J., Beltz, C., Pellicio, A., Powell, M., & Vale, W. (2003). Qualitative research on
adolescent risk using email: A methodological assessment. Qualitative Sociology, 26(1), 111–124.

Hertz, R., & Imber, J. B. (Eds.). (1995). Studying elites using qualitative methods. Thousand Oaks, CA:
Sage.

Hines, L. (2011). The treatment views and recommendations of substance-abusing women: A meta-synthesis.
Qualitative Social Work, 12, 473–479.

Hoagwood, K., & Olin, S. (2002). The NIMH blueprint for change report: Research priorities in child and
adolescent mental health. Journal of the American Academy of Child and Adolescent Psychiatry, 41(7),
760–767.

Hochschild, A. (1983). The managed heart: The commercialization of human feeling. Berkeley, CA:

401
 University of California Press.

Hohmann, A., & Shear, M. K. (2002). Community-based intervention research: Coping with the “noise” of
real life in study design. American Journal of Psychiatry, 159, 201–207.

Holliday, A. (2007). Doing and writing qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Holstein, J. A., & Gubrium, J. F. (1995). The active interview. Thousand Oaks, CA: Sage.

Holstein, J. A., & Gubrium, J. F. (Eds.). (2007). Handbook of constructionist research. New York, NY:
Guilford Press.

Hong, Y., Mitchell, S. G., Peterson , J. A., Latkin, C. A., Tobin, K., & Gann, D. (2005). Ethnographic
process evaluation: Piloting an HIV prevention program among injection drug users. International Journal
of Qualitative Methods, 4(1), Article 1.

Houston, S., & Mullan-Jensen, C. (2011). Towards depth and width in qualitative social work: Aligning
interpretive phenomenological analysis with the theory of social domains. Qualitative Social Work, 11,
266–281.

Hsieh, H., & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative Health
Research, 15(9), 1277–1288.

Humphrey, C. (2012). Dilemmas in doing insider research in professional education. Qualitative Social
Work, 12, 572–586.

Humphries, L. (1970). Tearoom trade: Impersonal sex in public places. Chicago, IL: Aldine.

Hyden, M., & Overlien, C. (2004). “Doing” narrative analysis. In D. K. Padgett (Ed.), The qualitative
research experience (pp. 250–268). Pacific Grove, CA: Thomson Learning.

Illingworth, N. (2001). The Internet matters: Exploring the use of the Internet as a research tool. Sociological
Research On-Line, 6(2), U96–U112. Retrieved from http://www.socresonline.org.uk/6/2/illingworth.html

Inui, T. S., & Frankel, R. M. (1991). Evaluating the quality of qualitative research. Journal of General

402
 Internal Medicine, 6, 485–487.

Israel, B. A., Eng, E., Schulz, A. J., & Parker. E. A. (Eds.). (2005). Methods in community-based
participatory research for health. San Francisco, CA: Jossey-Bass.

Iversen, R. R. (2008). “Getting out” in ethnography: A seldom-told story. Qualitative Social Work,
8(1),9–26.

Iversen, R., & Armstrong, A. L. (2006). Jobs aren’t enough: Toward a new economic mobility for low-
income families. Philadelphia, PA: Temple University Press.

Janesick, V. J. (2000). The choreography of qualitative research designs: Minuets, improvisations and
crystallization. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 379–400).
Thousand Oaks, CA: Sage.

Johnson, B., & Clarke, J. (2003). Collecting sensitive data: The impact on the researcher, Qualitative Health
Research 13, 421–434.

Johnson, P. J., & Onwuegbuzie, J. A. (2004). Mixed methods research: A research paradigm whose time has
come. Educational Researcher, 33(7), 14–26.

Johnstone, P. L. (2004). Mixed methods, mixed methodology health services research in practice. Qualitative
Health Research, 14(2), 259–271.

Jones, L., & Wells, K. (2007). Strategies for academic and clinician engagement in community-based
partnered research. Journal of the American Medical Association, 297(4), 407–410.

Jones, S. H., Adams, T. E., & Ellis, C. (Eds.). (2013). Handbook of autoethnography. Walnut Creek, CA:
Left Coast Press.

Kaptchuk, T. J., Shaw, J., Kerr, C. E., Conboy, L. A., Kelley, J. M., Csordas, T. J., Lembo, A. J., & Jacobson
E. E. (2009). “Maybe I made up the whole thing’’: Placebos and patients’ experiences in a randomized
controlled trial. Culture, Medicine and Psychiatry, 33, 382–411.

Katz, A., & Mishler, E.G. (2003). Close encounters: Exemplars of process-oriented qualitative research in
health care. Qualitative Research, 3, 35–56.

403
Kelle, U., & Erzberger, C. (2004). Qualitative and quantitative methods: Not in opposition. In U. Flick, E.
von Kardorff, & I. Steinke, (Eds.), A companion to qualitative research (pp. 172–175). London, UK: Sage.

Kelling, G. L., & Coles, C. M. (1996). Fixing broken windows: Restoring social order and reducing crime in
our communities. New York, NY: The Free Press.

Kerner, J. F., Yedidia, M., Padgett, D., Muth, B., Washington, K. S., Tefft, M., Yabroff, K., R., Makariou,
E., Freeman, H., & Mandelblatt, J. S. (2003). Realizing the promise of breast cancer screening: Clinical
follow-up after abnormal screening among Black women. Preventive Medicine, 37, 92–101.

Khantzian, E. J. (1985). The self-medication hypothesis of addictive disorders: Focus on heroin and cocaine
dependence. American Journal of Psychiatry, 142, 1259–1264.

Kim, K. H., Yoon, H. W., & Park, H. W. (2004). Spatiotemporal brain activation pattern during
word/picture perception by native Koreans. NeuroReport, 15, 1099–1103.

Kincheloe, J. L., & McLaren, P. (2000). Rethinking critical theory and qualitative research. In N. K. Denzin
& Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed., pp. 279–314). Thousand Oaks, CA:
Sage.

Klinenberg, E. (2002). Heat wave: A social autopsy of disaster in Chicago. Chicago, IL: University of
Chicago Press.

Kosinki, M., Matz, S., Gosling, S. D., Popov, V., & Stillwell, D. (2015). Facebook as a research tool for the
social sciences: Opportunities, challenges, ethical considerations, and practical guidelines. American
Psychologist, 70, 543–556.

Kotkin, S. (2002, September 7). A world war among professors: A clash between number crunchers and
specialists in a single region. The New York Times, B9–B11.

Kranke, D., Jackson, S. E., Taylor, D. A., Languth, J., & Floersch, J. (2015). “I’m loving life”: Adolescents’
experiences of living with a mental illness. Qualitative Social Work, 14, 102–118.

Krueger, R. A. (1994). Focus groups: A practical guide for applied research (2nd ed.). Thousand Oaks, CA:
Sage.

404
Kusenbach, M. (2003). Street phenomenology: The go-along as ethnographic research tool. Ethnography, 4,
455–485.

Labov, W., & Waletzky, J. (1967). Narrative analysis: Oral versions of personal experience. In J. Helm (Ed.),
Essays on the verbal and visual arts (pp. 12–44). Seattle: University of Washington Press.

Ladson-Billings, G. (2000). Racialized discourses and ethnic epistemologies. In N. K. Denzin & Y. S. (Eds.),
Handbook of qualitative research (pp. 258–278). Thousand Oaks, CA: Sage.

Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago, IL: University of Chicago Press.

Laing, R. D. (1965). The divided self: An existential study in sanity and madness. London, UK: Pelican
Press.

Langellier, K. M., & Peterson, E. E. (2004). Performing narrative: Storytelling in daily life. Philadelphia, PA:
Temple University Press.

LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces, 13, 230–237.

Leamy, M., Bird, V., LeBoutillier, C., Williams, J., & Slade, M. (2011). Conceptual framework for personal
recovery in mental health: Systematic review and narrative synthesis. British Journal of Psychiatry, 199,
445–452.

LeClerc, C. M., & Kensinger, E. A. (2011). Neural processing of emotional pictures and words: A
comparison of young and older adults. Developmental Neuropsychology, 36, 519–538.

LeCompte, M. D., & Schensul, J. J. (1999). Analyzing and interpreting ethnographic data. Walnut Creek,
CA: Altamira Press.

Levin, M., & Greenwood, D. (2001). Pragmatic action research and the struggle to transform universities
into learning communities. In P. Reason & H. Bradbury (Eds.), Handbook of action research (pp.
103–113). London, UK: Sage.

Levy, R. I., & Hollan, D. W. (2000). Person-centered interviewing and observation. In H. R. Bernard (Ed.),

405
 Handbook of methods in cultural anthropology (pp. 333–364). Walnut Creek, CA: Altamira Press.

Lewin, K. (1946). Action research and minority problems. Journal of Social Issues, 4, 34–46.

Liebow, E. (1967). Talley’s corner: A study of Negro street corner men. Boston, MA: Little, Brown.

Liebow, E. (1993). Tell them who I am: The lives of homeless women. New York, NY: Penguin.

Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage.

Lincoln, Y. S., & Guba, E. G. (2013). The constructivist credo. Walnut Creek, CA: Left Coast Press.

Littell, J. H., Corcoran, J., & Pillai, V. (2008). Systematic reviews and meta-analysis. New York, NY: Oxford
University Press.

Lofland, J. (2002). Analytic ethnography. In A. M. Huberman & M. B. Miles (Eds.), The qualitative
researcher’s companion (pp. 137–170). Thousand Oaks, CA: Sage.

Lofland, J., & Lofland, L. (1995). Analyzing social settings: A guide to qualitative observation and analysis
(3rd ed.). Belmont, CA: Wadsworth.

Lofland, J., & Snow, D. A. (2005). Analyzing social settings: A guide to qualitative observation and analysis.
Belmont, CA: Wadsworth.

Longhorn, J., & Floersch, J. (2012). The coming crisis in social work: Some thoughts on social work and
science. Research on Social Work Practice, 22, 499–519.

Longhofer, J., Floersch, J., & Hoy, J. (2013). Qualitative methods for practice research. New York, NY:
Oxford University Press.

Lopez, K. A., & Willis D. G., (2004). Descriptive versus interpretive phenomenology: Their contributions to
nursing knowledge. Qualitative Health Research, 14, 726–735.

Lubet, S. (2015). Ethics on the run: New rambler review. Chicago, IL: University of Chicago Law School.

406
Luders, C. (2004). Field observation and ethnography. In U. Flick, E. von Kardorff, & I. Steinke (Eds.), A
companion to qualitative research (pp. 222–230). London, UK: Sage.

Lynd, R. S., & Lynd, H. M. (1937). Middletown in transition: A study in cultural conflicts. New York, NY:
Harcourt Brace.

Lynd, R. S., & Lynd, H. M. (1956). Middletown: A study in modern American culture. New York, NY:
Harcourt Brace.

MacClean, L. M., Meyer, M., & Estable, A. (2004). Improving accuracy of transcripts in qualitative research.
Qualitative Health Research, 14(1), 113–123.

MacGregor, T. E., Rodger, S., Cummings, A. L., & Leschied, A. W. (2006). The needs of foster parents: A
qualitative study of motivation, support and retention. Qualitative Social Work, 5(3), 351–368.

Macia, L. (2015). Using clustering as a tool: Mixed methods in qualitative data analysis. The Qualitative
Report, 20, 1083–1094.

Madison, D. S. (2005). Critical ethnography: Methods, ethics, and performance. Thousand Oaks, CA: Sage.

Major, C., & Savin-Baden, M. (2010). An introduction to qualitative research synthesis: Managing the
information explosion in social science research. New York, NY: Routledge.

Malinowski, B. (1922). Argonauts of the Western Pacific. London, UK: Routledge.

Malone, R. E., Yerger, V. E., McGruder, C., & Froelicher, E. (2006). “It’s like Tuskegee in reverse”: A case
study of ethical tensions in institutional review board review of community-based participatory research.
American Journal of Public Health, 96, 1914–1919.

Mancini, M. A. (2005). “Making sense of it all”: Consumer providers’ theories about factors facilitating and
impeding recovery from psychiatric disabilities. Psychiatric Rehabilitation Journal, 29(1), 48–55.

Manderson, L., & Aaby, P. (1992). An epidemic in the field? Rapid assessment procedures and health
research. Social Science & Medicine, 35, 839–850.

407
Manderson, L., Bennett, E., & Andajani-Sutjaho, S. (2006). The social dynamics of the interview: Age, class
and gender. Qualitative Health Research, 16(10), 1317–1334.

Manderson, L., Kelaher, M., & Woelz-Stirling, N. (2001). Developing qualitative databases for multiple
users. Qualitative Health Research, 11(2), 149–160.

Manicas, P. T., & Secord, P. F. (1982). Implications for psychology of the new philosophy of science.
American Psychologist, 38, 390–413.

Margolis, E., & Pauwels, L. (Eds.). (2011). The Sage handbook of visual research methods. Thousand Oaks,
CA: Sage.

Markham, A. N. (2005). The methods, politics, and ethics of representation in online ethnography. In N. K.
Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (3rd ed., pp. 793–820). Thousand Oaks,
CA: Sage.

Marrozzo, J. M., Ramjee, G., Richardson, B. Gomez, K., Mogdi, N, Nair, G. . . . Voice Study Team. (2015).
Tenofovir pre-exposure prophylaxis for HIV infection among African women. New England Journal of
Medicine, 372, 509–518.

Marshall, C., & Rossman, G. B. (2010). Designing qualitative research (5th ed.). Thousand Oaks, CA: Sage.

Mattingly, C. (1998). Healing dramas and clinical plots: The narrative structure of experience. Cambridge
UK: Cambridge University Press.

Mattingly, C. (2010). The paradox of hope: Journeys through a clinical borderland. Berkeley, CA: University
of California Press.

Maxwell, J. (2002). Understanding and validity in qualitative research. In A. M. Huberman & M. B. Miles
(Eds.), The qualitative researcher’s companion (pp. 37–62). Thousand Oaks, CA: Sage.

Maxwell, J. A. (2010). Using numbers in qualitative research. Qualitative Inquiry, 16, 475–482.

Maxwell, J. A. (2012). Qualitative research design: An interactive approach (3rd ed.). Thousand Oaks, CA:
Sage.

408
Mayring, P. (2004). Qualitative content analysis. In U. Flick, E. von Kardorff, & I. Steinke (Eds.), A
companion to qualitative research (pp. 266–270). London, UK: Sage.

Mays, N., & Pope, C. (2000). Qualitative research in health care: assessing quality in qualitative research.
British Medical Journal, 320, 50–52.

McCall, M. (2000). Performance ethnography: A brief history and some advice. In N. K. Denzin & Y. S.
Lincoln (Eds.), Handbook of qualitative research (pp. 421–434). Thousand Oaks, CA: Sage.

McCracken, G. (1988). The long interview. Newbury Park, CA: Sage.

Mead, M. (1928). Coming of age in Samoa. New York, NY: New American Library.

Menand, L. (2001). The metaphysical club: A story of ideas in America. New York, NY: Farrar, Straus, &
Giroux.

Mendlinger, S., & Cwikel, J. (2008). Spiraling between qualitative and quantitative data on women’s health
behaviors: A double helix model for mixed methods. Qualitative Health Research, 18, 280–293.

Merriam, S. B. (1998). Qualitative research and case study applications in education. San Francisco, CA:
Jossey-Bass.

Mertens, D. (2015). Mixed methods and wicked problems. Journal of Mixed Methods Research, 9, 3-6.

Merton, R. K., Fiske, M., & Kendall, P. (1956). The focused interview. Glencoe, IL: Free Press.

Metcalfe, F., & Humphries, C. (2002). Fostering action research and action research in fostering. Qualitative
Social Work, 1, 435–450.

Miles, M. B., & Huberman, A. M. (Eds.). (1994). Qualitative data analysis: An expanded sourcebook (2nd
ed.). Thousand Oaks, CA: Sage.

Miller, D., & Slater, D. (2000). The Internet: An ethnographic approach. New York, NY: Berg.

Miller, J., & Glassner, B. (2011). The inside and the outside: Finding realities in interviews. In D. Silverman

409
 (Ed.). Qualitative research: Issues of theory, method and practice (3rd ed.), pp. 291–309. London, UK:
Sage.

Mills, A. J., Derepos, G., & Wiebe, E. (2010). Encylopedia of case study research. Thousand Oaks, CA:
Sage.

Mills, J., Bonner, A., & Francis, K. (2006). The development of constructivist grounded theory. International
Journal of Qualitative Methods, 5(1), Article 3.

Minkler, M., & Wallerstein, N. (2003). Community-based participatory research for health. San Francisco,
CA: Jossey-Bass.

Mishler, E. (1986). Research interviewing: Context and narrative. Cambridge, MA: Harvard University Press.

Morgan, D. L. (1993). Qualitative content analysis: A guide to paths not taken. Qualitative Health Research,
1, 112–121.

Morgan, D. L. (1997). Focus groups as qualitative research. Thousand Oaks, CA: Sage.

Morgan, D. L. (2014). Pragmatism as a paradigm for social research. Qualitative Inquiry, 20, 1045–1053.

Morrow, S. (2005). Quality and trustworthiness in qualitative research in counseling psychology. Journal of
Counseling Psychology, 52(2), 250–260.

Morrow, S. L., & Smith, M. L. (1995). Constructions of survival and coping by women who have survived
childhood sexual abuse. Journal of Counseling Psychology, 42(1), 24–33.

Morse, J. M. (1985). Nursing research: The application of qualitative approaches. London, UK: Blackwell.

Morse, J. M. (1991). Approaches to qualitative-quantitative methodological triangulation. Nursing Research,

40, 120–123.

Morse, J. M. (1994). Designing funded qualitative research. In N. K. Denzin & Y. L. Lincoln (Eds.),
Handbook of qualitative research (pp. 220–235). Thousand Oaks, CA: Sage.

410
Morse, J. M. (1995). The significance of saturation. Qualitative Health Research, 5, 147–149.

Morse, J. (2005). Evolving trends in qualitative research: Advances in mixed-method design. Qualitative
Health Research, 15(5), 583–585.

Morse, J. M. (2006). The politics of evidence. Qualitative Health Research, 16(3), 395–404.

Morse, J. M. (online August, 2015). Critical analysis of strategies for determining rigor in qualitative inquiry.
Qualitative Health Research.

Moustakas, C. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage.

Mufti, G., Towell, T., & Cartright, T. (2015). Pakistani children’s experiences of growing up with beta-
thalassemia major. Qualitative Health Research, 25, 286–396.

Munhall, P. L. (1994). Qualitative research proposals and reports: A guide. New York, NY: National League
for Nursing Press.

Munhall, P. L. (2012). Nursing research. Sudbury, MA: Jones and Bartlett Learning.

Nelson, G., Ochocka, J., Griffin, K., & Lord, J. (1998). “Nothing about me, without me”: Participatory action
research with self-help/mutual aid organizations. American Journal of Community Psychology, 26(6),
881–913.

Newman, K., Fox, C., Roth, W., & Mehta, J. (2004). Rampage: The social roots of school shootings. New
York, NY: Basic Books.

Neyfakh, L. (June 18, 2015). The ethics of ethnography. Slate. Retrieved from
http://www.slate.com/articles/news_and_politics/crime/2015/06/alice_goffman_s_on_the_run_is_the_sociologist_to_blame_f

Noblit, G. W., & Hare, D. R. (1988). Meta-ethnography: Synthesizing qualitative studies. Newbury Park,
CA: Sage.

Noth, W. (2011). Visual semiotics: Key features and an application to picture ads. In E. Margolis & L.
Pauwels (Eds.), The Sage handbook of visual research methods (pp. 298–316). Thousand Oaks, CA: Sage.

411
Noyes, J., Popay, J., Pearson, A., Hannes, K., & Booth, A. (2008). Qualitative research and Cochrane
reviews. In J. Higgins & S. Green (Eds.), Cochrane handbook for systematic reviews of interviews (pp.
101–114). London, UK: Wiley & Sons.

Nussbaum, M. (2000). Women and human development: The capabilities approach. Cambridge, MA:
Cambridge University Press.

O’Brien, B. C., Harris, I. B., Beckman, T. J., Reed, D. A., & Cook, D.A. (2014). Standards for reporting
qualitative research: A synthesis of recommendations. Academic Medicine, 89, 1245–1251.

Oktay, J. (2012). Grounded theory. New York, NY: Oxford University Press.

Olesen, V. L. (2000). Feminisms and qualitative research at and into the millennium. In N. K. Denzin & Y.
S. Lincoln (Eds.), Handbook of qualitative research (pp. 215–256). Thousand Oaks, CA: Sage.

Onwuegbuzie, A. (2012). Putting the mixed back into quantitative and qualitative research in educational
research and beyond: Moving towards the “radical middle.” International Journal of Multiple Research
Approaches, 6, 192–219.

Padgett, D. K. (1998). Does the glove really fit? Qualitative research and clinical social work practice. Social
Work, 43, 373–381.

Padgett, D. K. (Ed.). (2004a). The qualitative research experience. Belmont, CA: Thomson.

Padgett, D. K. (2004b). Spreading the word: Writing up and disseminating qualitative research. In D. K.
Padgett (Ed.), The qualitative research experience (pp. 285–296). Belmont, CA: Thomson.

Padgett, D. K. (2007). There’s no place like (a) home: Ontological security in the third decade of the
homelessness crisis. Social Science & Medicine, 64, 1925–1936.

Padgett, D. K. (2009). Qualitative and mixed methods in social work knowledge development. Social Work,
54, 101–105.

Padgett, D. K. (2015). Qualitative methods in evaluation. In D. Royse, B. T. Thyer, & D. K. Padgett,

Program evaluation (6th ed., pp. 48–61). Pacific Grove, CA: Cengage.

412
Padgett, D. K., Gulcur, L., & Tsemberis, S. (2006). Housing first services for the psychiatrically disabled
homeless with co-occurring substance abuse. Research on Social Work Practice, 16, 74–83.

Padgett, D. K., Hawkins, R. L., Abrams, C., & Davis, A. (2006). In their own words: Trauma and substance
abuse in the lives of formerly homeless women with serious mental illness. American Journal of
Orthopsychiatry, 76(1), 461–467.

Padgett, D. K., & Henwood, B.F. (2009). Obtaining large-scale funding forempowerment-oriented
qualitative research: A report from personal experience. Qualitative Health Research, 19, 868–875

Padgett, D. K., Henwood, B. F., Abrams, C., & Davis, A. (2008). Engagement and retention in care among
formerly homeless adults with serious mental illness: Voices from the margins. Psychiatric Rehabilitation
Journal, 31(3), 226–233.

Padgett, D. K., Mathew, R., & Conte, S. (2004). Peer debriefing and support groups. In D. K. Padgett (Ed.),
The qualitative research experience (pp. 156–169). Pacific Grove, CA: Thomson Learning.

Padgett, D. K., Patrick, C., Burns, B. J., & Schlesinger, H. J. (1994). Ethnicity and use of outpatient mental
health services in a national insured population. American Journal of Public Health, 84, 222–226.

Padgett, D. K, Smith, B.T., Choy-Brown, M., Tiderington, E., & Mercado, M. (February 16, 2016, online
first). Trajectories of recovery among formerly homeless adults with serious mental illness. Psychiatric
Services.

Padgett, D. K., Stanhope, V., Henwood, B. F., & Stefancic, A. (2011). Substance use outcomes among
homeless clients with serious mental illness: Comparing Housing First with Treatment First programs.
Community Mental Health Journal, 47, 227–232.

Padgett, D. K., Tran Smith, B., Choy-Brown, M., Tiderington, D., & Mercado, M. (2016). Trajectories of
recovery among formerly homeless adults with serious mental illness. Psychiatric Services.

Padgett, D. K, Tran Smith, B., Derejko, K., Henwood, B. F., & Tiderington, E. (2013). A picture is worth . .
. ? Using individual photo-elicitation to enhance interviews with vulnerable populations. Qualitative Health
Research. 23, 1435–1444.

Padgett, D. K., Tran Smith, B., Henwood, B. F., & Tiderington, E. (2012). Life course adversity in the lives

413
 of formerly homeless persons with serious mental illness: Context and meaning. American Journal of
Orthopsychiatry, 82, 421–430.

Padgett, D. K., Yedidia, M., Kerner, J., & Mandelblatt, J. (2001). The emotional consequences of false
positive mammography: African-American women’s reactions in their own words. Women and Health, 33,
1–14.

Palinkas, L. A., Aarons, G. A., Horvitz, S., Chamberlain, P., Hurlburt, M., & Landsverk, J. (2011). Mixed
methods designs in implementation research. Administration and Policy in Mental Health, 38, 44–53.

Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

Pederson, J. R. (2013). Disruptions of individual and cultural identities. Narrative Inquiry, 23, 302–322.

Peirce, C. (1934). Collected papers of Charles Sanders Peirce. Vol. 5, pragmatism and pragmaticism, edited
by C. Hartshorne and P. Weiss. Cambridge, MA: Harvard University Press.

Pelto, P. (2015). What is so new about mixed methods? Qualitative Health Research, 25, 734–745.

Perreault, M., Pawliuk, N., Veilleux, R., & Rousseau, M. (2006). Qualitative assessment of mental health
service satisfaction: Strengths and limitations of a self-administered measure. Community Mental Health
Journal, 42(3), 233–242.

Petryna, A. (2009). When experiments travel: Clinical trials and the global search for human subjects.
Princeton, NJ: Princeton University Press.

Pink, S. (2012). Advances in visual methodology. London: Sage.

Pinto, R. M., Spector, A. Y., & Valera, P. A. (2011). Exploring group dynamics for integrating scientific and
experiential knowledge in community advisory boards for HIV research, AIDS Care, 23, 1006–1013.

Polkinghorne, D. E. (1988). Narrative knowing and the human sciences. Albany: State University of New
York Press.

Popay J., Williams G., & Rogers A. (1998). Rationale and standards for the systematic review of qualitative

414
 literature in health services research. Qualitative Health Research, 8, 341–351.

Powdermaker, H. (1966). Stranger and friend: The way of an anthropologist. New York, NY: W. W. Norton
.

Proctor, E., Knudsen, K. J., Fedoracivius, N., Hovmand, P., Rosen, A., & Perron, B. (2007). Implementation
of evidence-based practice in community behavioral health. Administration and Policy in Mental Health,
44, 479–488.

Rabinow, P., & Sullivan, W. M. (Eds.). (1979). Interpretive social science: A reader. Berkeley: University of
California Press.

Ragin, C. C. (1987). The comparative method. Moving beyond qualitative and quantitative strategies.
Berkeley: University of California Press.

Rallis, S. F., & Rossman, G. B. (2003). Mixed methods in evaluation contexts: A pragmatic framework. In A.
Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in social and behavioral research (pp.
491–512). Thousand Oaks, CA: Sage.

Reason, P., & Bradbury, H. (2008). Handbook of action research (2nd ed.). Thousand Oaks, CA: Sage.

Reason, P., & Bradbury-Huang, H. (2013). The Sage handbook of action research. Thousand Oaks, CA:
Sage.

Reinharz, R. (1992). Feminist methods in social research. New York, NY: Oxford University Press.

Rice, E., & Karnik, N. S. (2012). Network science and social media. Journal of the American Academy of
Child & Adolescent Psychiatry, 51(6), 563–5.

Rice, E., Lee, A. W., & Taitt, S. (2011). Cell phone use among homeless youth: Potential for new health
interventions and research. Journal of Urban Health, 88, 1175–82.

Richardson, L. (2000). Writing: A method of inquiry. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of
qualitative research (2nd ed., pp. 923–938). Thousand Oaks, CA: Sage.

415
Riessman, C. K. (1990). Divorce talk: Women and men make sense of personal relationships. Rutgers, NJ:
Rutgers University Press.

Riessman, C. K. (2008). Narrative methods for the human sciences. Thousand Oaks, CA: Sage.

Riessman, C. K., & Quinney, L. (2005). Narrative in social work: A critical review. Qualitative Social Work,
4(4), 391–412.

Rodriquez, J. (2013). Narrating dementia: Self and community in an on-line forum. Qualitative Health
Research, 23, 1215–1227.

Rolfe, G. (2006). Validity, trustworthiness and rigour: Quality and the idea of qualitative research. Journal of
Advanced Nursing, 53(3), 304–310.

Rorty, R. (1998). Truth and progress: Philosophical papers III. Cambridge, MA: Cambridge University
Press.

Rosaldo, R. (1989). Culture and truth: The remaking of social analysis. Boston, MA: Beacon.

Rose, G. (2007). Visual methodologies: An introduction to the interpretation of visual materials. London,
UK: Sage.

Ruckdeschel, R., Earnshaw, P., & Firrek, A. (1994). The qualitative case study and evaluation: Issues,
methods, and examples. In E. Sherman & W. J. Reid (Eds.), Qualitative research in social work (pp.
251–264). New York, NY: Columbia University Press.

Rugg, G., & McGeorge, P. (2005). The sorting techniques: A tutorial paper on card sorts, picture sorts, and
item sorts. Expert Systems, 22, 94–107.

Ryan, G. W., & Bernard, H. R. (2000). Data management and analysis methods. In N. K. Denzin and Y. S.
Lincoln (Eds.), Handbook of qualitative research (pp. 769–802). Thousand Oaks, CA: Sage.

Ryan, G. W., & Bernard, H. R. (2003). Techniques to identify themes. Field Methods, 15(1), 85–109.

Sacks, H., & Garfinkel, H. (1970). On formal structures of practical action. In J. C. McKinney & E. A.

416
 Tiryakian (Eds.), Theoretical sociology (pp. 338–366). New York, NY: Appleton-Century-Crofts.

Saldana, J. (2003). Longitudinal qualitative research. Walnut Creek, CA: Altamira.

Saldana, J. (2015). The coding manual for qualitative researchers (3rd ed.). Thousand Oaks, CA: Sage.

Saldana, J. (2016). The coding manual for qualitative researchers (3rd ed.). Thousand Oaks, CA: Sage.

Saleebey, D. (2005). The strengths perspective in social work practice (4th ed.). Boston, MA: Allyn & Bacon.

Salganik, M. J., & Heckathorn, D. D. (2004). Sampling and estimation in hidden populations using
respondent-driven sampling. Sociological Methodology, 34, 193–239.

Sandelowski, M. (1993). Rigor, or rigor mortis: The problem of rigor in qualitative research revisited.
Advances in Nursing Science, 16, 1–8.

Sandelowski, M. (2000). Combining qualitative and quantitative sampling, data collection, and analysis
techniques in mixed methods studies. Research in Nursing & Health, 23, 246–255.

Sandelowski, M., & Barroso, J. (2002). Reading qualitative studies. International Journal of Qualitative
Methods, 1(1), Article 5.

Sandelowski, M., & Barroso, J. (2003). Classifying the findings in qualitative studies. Qualitative Health
Research, 13(7), 905–923.

Sandelowski, M., Docherty, R., & Emden, C. (1997). Qualitative meta-synthesis: Issues and techniques.
Research in Nursing & Health, 20, 365–371.

Sandelowski, M., & Jones, L. C. (1995). “Healing fictions”: Stories of choosing in the aftermath of the
detection of fetal anomalies. Social Science and Medicine, 42, 353–361.

Sands, R. G. (2004). Narrative analysis: A feminist approach. In D. K. Padgett (Ed.), The qualitative research
experience (pp. 48–62). Belmont, CA: Thomson.

Sandstrom, B., Willman, A., Svensson, B., & Borglin, G. (2015). Perceptions of national guidelines and their

417
 (non)implementation in mental health care: A deductive and inductive content analysis. Implementation
Science, 10, 43–58.

Schwandt, T. A. (1994). Constructivist, interpretivist approaches to human inquiry. In N. K. Denzin & Y. S.

Lincoln (Eds.), Handbook of qualitative research (pp. 118–137). Thousand Oaks, CA: Sage.

Scrimshaw, S. C., Carballo, M., Ramos, L., & Blair, B. A. (1991). The AIDS Rapid Anthropological
Assessment Procedures: A tool for health education planning and evaluation. Health Education Quarterly,
18(1), 111–123.

Scriven, M. (1967). The methodology of evaluation. AERA Monograph Series in Curriculum Evaluation, 1,
39–83.

Seale, C. (2002). Qualitative issues in qualitative inquiry. Qualitative Social Work 1, 97–110.

Seale, C. (2004). Researching society and culture. London, UK: Sage.

Seale, C., Charteris-Black, J., McFarlane, A., & McPherson, A. (2010). Interviews and Internet forums: A
comparison of two sources of qualitative data. Qualitative Health Research, 20, 595–606.

Seale, C., & Silverman, S. (1997). Ensuring rigour in qualitative research. European Journal of Public Health,
7, 379–384.

Seidman, I. (2006). Interviewing as qualitative research. New York, NY: Teachers College Press.

Sen, A. (1999). Development as freedom. New York, NY: Knopf.

Shadish, W., Cook, T., & Campbell, D. (2002). Experimental and quasi-experimental designs for generalized
causal inference. Boston, MA: Houghton Mifflin.

Shibusawa, T., & Lukens, E. (2004). Analyzing qualitative data in a cross-language context: A collaborative
model. In D. K. Padgett (Ed.), The qualitative research experience (pp. 175–186). Belmont, CA:
Thomson.

Shibusawa, T., & Padgett, D. K. (2009). Out of sync: A life course perspective on aging among formerly

418
 homeless adults with serious mental illness. Journal of Aging Studies, 23(3), 188–196

Shore, N., Wong, K. A., Seifer, S. D., Grignon, J., & Gamble, V. N. (2008). Advancing the ethics of
community-based participatory research. Journal of Empirical Research on Human Research Ethics, 10,
1–4.

Sidnell, J., & Stivers, T. (Eds.). (2013). The handbook of conversation analysis. Oxford, UK: Blackwell Press.

Silverman, D. (2006). Interpreting qualitative data (3rd ed.). Thousand Oaks, CA: Sage.

Singer, B., Ryff, C. D., Carr, B., & Magee, M. J. (1998). Linking life histories and mental health: A person-
centered approach. Sociological Methodology, 28, 1–51.

Smith, B. & Sparkes, A. C. (2006). Narrative inquiry in psychology: Exploring the tensions within.
Qualitative research in psychology. 3, 169–192.

Smith, D. E. (2005). Institutional ethnography: A sociology for the people. Lanham, MD: Altamira Press.

Smith, J. A. (1996). Beyond the divide between cognition and discourse: Using interpretive phenomenological
analysis in health psychology. Psychology and Health, 11, 13–24.

Smith, J. A., & Osborn, M. (2009). Interpretive phenomenological analysis. In J. A. Smith, P. Flowers, & M.
Larkin (Eds.). Interpretive phenomenological analysis: Theory, method and research (pp. 53–80). London,
UK: Sage.

Snow, D. A., & Anderson, L. (1987). Identity work among the homeless: The verbal construction and avowal
of personal identities. American Journal of Sociology, 92, 1336–1371.

Snow, D. A., & Anderson, L. (1991). Researching the homeless: The characteristic features and virtues of the
case study. In J. R. Feagin, A. M. Orum, & G. Sjoberg (Eds.), A case for the case study (pp. 148–173).
Chapel Hill: University of North Carolina Press.

Snow, D. A., Morrill, C., & Anderson, L. (2003). Elaborating analytic ethnography: Linking fieldwork and
theory. Ethnography 4, 181–200.

419
Spalding, N. J., & Phillips, T. (2007). Exploring the use of vignettes: From validity to trustworthiness.
Qualitative Health Research, 17, 954–962.

Speer, S. A. (2002). Natural and contrived data: A sustainable distinction? Discourse Studies, 4, 511–525.

Spencer, R. (2006). Understanding the mentoring process between adolescents and adults. Youth & Society,
37, 287–315.

Spradley, J. P. (1979). The ethnographic interview. New York, NY: Holt, Rinehart, & Winston.

Stack, C. B. (1974). All our kin: Strategies for survival in a black community. New York, NY: Harper
Colophon.

Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.

Stake, R. E. (2005). Multiple case study analysis. New York, NY: Guilford Press.

Staller, K. (2002). Musings of a skeptical software junkie and the HYPER-RESEARCH fix. Qualitative
Social Work, 1, 473–487.

Staller, K. (2012). Epistemological boot camp: What every qualitative researcher needs to know to survive in
the academy. Qualitative Social Work, 12, 395–413.

Stange, K. C., Miller, W. L., Crabtree, B. F., O’Connor, P. J., & Zyzanski, S. J. (1994). Integrating
qualitative and quantitative research methods. Family Medicine, 21, 448–451.

Steinmetz, A. M. (1991). Doing. In M. Ely, M. Anzul, T. Friedman, D. Garner, & A. M. Steinmetz, Doing
qualitative research: Circles within circles (pp. 41–68). London, UK: Falmer.

Steinmetz, G. (Ed.). (2005). The politics of method in human sciences: Positivism and its epistemological
others. Durham, NC: Duke University Press.

Stoller, E. P., Webster, N. J., Blixen, C. E., McCormick, R. A., Hund, A. J., Perzynski, A. T., Kanuch, S.
W., Thomas, C. L., Kercher, K., & Dawson, N. V. (2009). Alcohol consumption decisions among
nonabusing drinkers diagnosed with hepatitis C: An exploratory sequential mixed methods study. Journal

420
 of Mixed Methods Research, 3, 65–86.

Strauss, A., & Corbin, J. (1994). Grounded theory methodology: An overview. In N. K. Denzin & Y. S.
Lincoln (Eds.), Handbook of qualitative research (pp. 273–285). Thousand Oaks, CA: Sage.

Stringer, E. (2013). Action research (4th ed.). Thousand Oaks, CA: Sage.

Susser, I. (March 5, 2015). Blame research design for failed HIV study. Al Jazeera. Retrieved from
http://america.aljazeera.com/opinions/2015/3/blame-research-design-for-failed-hiv-study.html September
13, 2015

Susser, M. (1997). Authors and authorship: Reform or abolition? American Journal of Public Health, 87,
1091–1092.

Szreter, S., & Woolcock, M. (2004). Health by association? Social capital, social theory, and the political
economy of public health. International Journal of Epidemiology, 33, 1–18.

Tandon, R. (1996). The historical roots and contemporary tendencies in participatory research: Implications
for health care. In K. de Koning & M. Martin (Eds.), Participatory research in health: Issues and
experiences (pp. 19–26). London, UK: Zed.

Tannen, D. (1990). You just don’t understand: Women and men in conversation. New York, NY: William
Morrow and Company.

Tannen, D. (2006). You’re wearing that? Understanding mothers and daughters in conversation. New York,
NY: Ballantine Books.

Tashakkori, A., & Creswell, J. W. (2007). The new era of mixed methods. Journal of Mixed Methods
Research, 1(1), 3–7.

Tashakkori, A., & Teddlie, C. (Eds.). (2010). Handbook of mixed methods in social and behavioral research
(2nd ed.). Thousand Oaks, CA: Sage.

Taylor, S. J. (1987). Observing abuse: Professional ethics and personal morality in field research. Qualitative
Sociology, 10, 288–302.

421
Taylor, S. J., & Bogdan, R. (1984). Introduction to qualitative research: The search for meanings (2nd ed.).
New York, NY: John Wiley.

Tedlock, B. (2000). Ethnographic and ethnographic representation. In N. K. Denzin & Y. S. Lincoln (Eds.),
Handbook of qualitative research (pp. 455–486). Thousand Oaks, CA: Sage.

ten Have, P. (2014). Doing conversation analysis (2nd ed). London, UK: Sage.

Teram, E., Schachter, C. L., & Stalker, C. A. (2005). The case for integrating grounded theory with
participatory action research: Empowering clients to inform professional practice. Qualitative Health
Research, 15(8), 1129–1140.

Tesch, R. (1990). Qualitative research: Analysis types and software tools. London, UK: Falmer.

Thomas, J., & Harden, A. (2008). Methods for the thematic synthesis of qualitative research in systematic
reviews. BMC Medical Research Methodology, 8, 45–52.

Thorne, S. (1998). Ethical and representational issues in qualitative secondary analysis. Qualitative Health
Research, 8(4), 547–555.

Thorne, S., Jensen, L., Kearny, M. H., Noblit, G., & Sandelowski, M. (2004). Qualitative metasynthesis:
Reflections on methodological orientation and ideological agenda. Qualitative Health Research, 14(10),
1342–1365.

Tiderington, E. (2015). The dilemmas of permanency and accountability: A qualitative investigation of

barriers to and facilitators of recovery-oriented practice in supportive housing. Dissertation retrieved from
ProQuest (10068).

Timmermans, S., & Tavory, I. (2012). Theory construction in qualitative research: From grounded theory to
abductive analysis. Sociological Theory, 30, 167–186.

Tjora, A. H. (2006). Writing small discoveries: An exploration of fresh observers’ observations. Qualitative
Research, 6(4), 429–451.

Tong, A., Lowe, A., Sainsbury, P., & Craig, J. (2008). Experiences of parents who have children with chronic
kidney disease: A systematic review of qualitative studies. Pediatrics, 121, 349–360.

422
Tong, A., Sainsbury, P., & Craig, J. (2007). Consolidated criteria for reporting qualitative research.
International Journal for Quality in Health Care, 19, 349–357.

Tracy, S. J. (2010). Qualitative quality: Eight “big tent” criteria for excellent qualitative research. Qualitative
Inquiry, 16, 837–851.

Tran Smith, B., Padgett, D. K., Choy Brown, M., & Henwood, B. F. (2015). Rebuilding lives and identities:
The role of place in mental health recovery. Health & Place, 33, 109–117.

Tsemberis, S., Gulcur, L., & Nakae, M. (2004). Housing First, consumer choice, and harm reduction for
homeless individuals with a dual diagnosis. American Journal of Public Health, 94(4), 651–656.

Twinn, S. (1997). An exploratory study examining the influence of translation on the validity and reliability of
qualitative data in nursing research. Journal of Advanced Nursing, 26, 418–423.

Ullman, S. (2014). Interviewing therapists about working with sexual assault survivors: Researcher and
therapist perspectives. Violence Against Women, 20, 1138–1146.

Ungar, M. (2006). “Too ambitious”: What happens when funders under-estimate the strength of qualitative
research design. Qualitative Social Work, 5(2), 261–277.

Vallée, J., Cadot, E., Roustit, C., Parizot, I., & Chauvin, P. (2011). The role of daily mobility in mental
health inequalities: The interactive influence of activity space and neighborhood of residence on depression.
Social Science & Medicine, 73, pp. 1133–1144.

van der Straten, A., Stadler J., Montgomery E., Hartmann, M., Magazi, B., et al. (2014). Women’s
experiences with oral and vaginal pre-exposure prophylaxis: The Voice-C qualitative study in Johannesburg,
South Africa, PLoS ONE 9(2): e89118.

Van Maanen, J. (1988). Tales of the field: On writing ethnography. Chicago, IL: University of Chicago Press.

van Manen, M. (1990). Researching lived experience. Albany: State University of New York Press.

van Manen, M. (Ed.). (2002). Writing in the dark: Phenomenological studies in interpretive inquiry. London,

423
 Canada: Althouse.

van Manen, M. (2006). Writing qualitatively, or the demands of writing. Qualitative Health Research, 16,
713–722.

Vaughan, D. (1996). The Challenger launch decision: Risky technology, culture and deviance at NASA.
Chicago, IL: University of Chicago Press.

Wackerbarth, S. B., Streams, M. E., & Smith, M. K. (2002). Capturing the insights of family caregivers:
Survey item generation with a coupled focus group / interview process. Qualitative Health Research, 12(8),
1141–1154.

Waldrop, D. (2004). Ethical issues in qualitative research with high-risk populations. In D. K. Padgett (Ed.),
The qualitative research experience (pp. 236–249). Belmont, CA: Thomson.

Walker, A. J., & Allen, K. R. (1991). Relationships between caregiving daughters and their elderly mothers.
The Gerontologist, 31, 389–396.

Walsh, D., & Downe, S. (2005). Meta-synthesis method for qualitative research: A literature review. Journal
of Advanced Nursing, 50, 204–211.

Wang, C., & Burris, M. A. (1997). Photovoice: Concept, methodology and use for participatory needs
assessment. Health Education & Behavior, 24(3), 369–387.

Wang, C. C., Morrel-Samuels, S., Hutchinson, P., Bell, L., & Pestronk, R. M. (2004). Flint photovoice:
Community building among youths, adults and policymakers. American Journal of Public Health, 94(6),
911–914.

Weiss, R. S. (1994). Learning from strangers: The art and method of qualitative interview studies. New York,
NY: Free Press.

Weller, S. C., & Romney, A. K. (1988). Systematic data collection. Newbury Park, CA: Sage.

Wells, K. (2011). Narrative inquiry. New York, NY: Oxford University Press.

424
White, M., & Epston, D. (1990). Narrative means to therapeutic ends. New York, NY: Norton.

Whyte, W. F. (1955). Street corner society (2nd ed.). Chicago, IL: University of Chicago Press.

Williams, C. C., & Collins, A. A. (2002). The social construction of disability in schizophrenia. Qualitative
Health Research, 12(3), 297–309.

Wilson, H. S., & Hutchison, S. (1991). Triangulation of qualitative methods: Heideggerian hermeneutics
and grounded theory. Qualitative Health Research, 1, 263–276.

Wimpenny, P., & Gass, J. (2000). Interviewing in phenomenology and grounded theory: Is there a difference?
Journal of Advanced Nursing, 31(6), 1485–1492.

Wolcott, H. F. (2001). Writing up qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Wolff, S. (2004). Analysis of documents and records. In U. Flick, E. von Kardoff, & I. Steinke (Eds.), A
companion to qualitative research (pp. 284–289). London, UK: Sage.

Yanos, P., & Hopper, K. (2008). On “false collusive objectification”: Becoming attuned to self-censorship and
interviewer biases in qualitative interviewing. International Journal of Social Research Methodology, 11,
229–237.

Yi-Frazier, J. P. Cochrane, K., Mitrovich, C., Pascual, M., Buscaino, E., Eaton, Panlasigui, N., Clopp, B., &
Malik, F. (2015). Using Instagram as a modified application of photovoice for storytelling and sharing in
adolescents with Type 1 diabetes. Qualitative Health Research, 25, 1372–1382

Yin, R. K. (2013). Case study research: Design and methods (5th ed.). Thousand Oaks, CA: Sage.

Zea, M. C., Aguilar-Pardo, M., Betancourt, F., Reisen, C. A., & Gonzalez, F. (2014). Mixed methods
research with internally displaced Colombian gay and bisexual men and transwomen. Journal of Mixed
Methods, 8, 212–221.

Zimmer, L. L. (2006). Qualitative meta-synthesis: A question of dialoguing with texts. Journal of Advanced
Nursing 53, 311–318.

425
Znaniecki, F. (1934). The method of sociology. New York, NY: Farrar & Rinehart.

426
Index

Abduction:
characteristics of, 13
interpretation stage, 188–189
Action research (AR):
characteristics of, 42–43
data analysis, 159–160
interpretation stage, 184, 199
mixed methods approach, 269, 270–271
research questions, 61
rigorous research, 221, 222
Addams, Jane, 8, 9
All Our Kin (Stack), 4
Amplified sampling, 177
Analogies, 194–195
Analysis-driven sampling techniques, 69
Analytic axes, 150, 157, 186, 187
Analytic ethnography, 13
Analytic expansion, 177
Analytic induction:
characteristics of, 12–13
interpretation stage, 188
Analytic triangulation, 216
A priori concepts:
in coding process, 164, 187, 200
in research design, 60
Archival materials, 133–134
Asylums (Goffman), 4, 244
ATLAS.ti, 142, 143–144
Attachment Theory, 11
Audio recording:
confidentiality, 84
interviews, 132
Audit trail:
for data analysis, 150
rigorous research, 217f, 220, 221
Authorship, 231, 235–238
Auto-coding, 143

427
Auto-ethnography, 46, 47
Axial coding, 199–200

Bias:
researcher, 213, 217f
respondent, 213–214, 217f
Big-tent approach, 209, 225–227
Budget justification, 284

Campbell Collaboration, 9–10, 202

Capabilities approach, 22, 189
Card sort technique, 121
Case study analysis:
characteristics of, 35–38
collective case study, 36, 155
data analysis, 155–156
instrumental case study, 36, 37, 155, 199
interpretation stage, 199
intrinsic case study, 36, 37, 155, 199
reports, 241
research questions, 61
research study, 38 (box)
rigorous research, 221
sample size, 70
Case summaries, 151
Case vignettes, 242
Categories, grounded theory (GT), 34
Certificate of Confidentiality (CoC), 84
Checklists, evaluating research, 223
Chicago School, 3, 5–6, 8–9, 33
Children:
ethical research, 81–82
interviews, 116–117
CINAHL Plus, 203–204 (box)
Clustering codes, 200
Coauthorship, 231, 235–238
Cochrane Collaboration, 9–10, 202
Coding:
approaches to, 163–164
auto-coding, 143
co-coding, 164

428
 codebook development, 170–175
code boundaries, 170
code labels, 169–170
code-worthy material, 164–169
data management log, 175–176
documentation procedures, 175–176
ethical research, 178
grounded theory (GT), 34, 163,177, 199–200
hybrid methods approach, 178–179
inductive coding, 34
inter-coder agreement, 164
interpretation stage, 199–200
in vivo codes, 170, 190, 191t, 195
meaning units, 164
memo writing, 176–177
mixed methods approach, 178–179
open coding, 34, 164–166, 167–168t
research study, 167–168t, 171–175
secondary analysis, 177–178
sensitizing concepts, 165
template approach, 165
See also Template Approach
theoretical codes, 170
Coercion of participant, 82–83
Collective case study, 36, 155
Common Rule guidelines, 86
Community-based participatory research (CBPR):
characteristics of, 42–44
data analysis, 159–160
ethics in, 91
Internet resources, 56
mixed methods approach, 270–271
modification influences, 47
research study, 44–45 (box)
rigorous research, 222
Community Campus Partnerships for Health (CCPH), 91
Community-engaged research (CER):
characteristics of, 43, 44
ethics in, 90–91
Complicated design approach, 262–263, 269–270

429
Computer-generated data, 133–134
Computer-generated interview, 130–132
Conceptual framework:
characteristics of, 11, 13–14
designing research, 60
Conceptualizing:
data analysis, 151–152
interpretation stage, 190, 191f
Concurrent design, 258, 260–262, 264
Confidentiality, 83–84, 88, 94
Confirmability, 210
Confirming sampling, 69
Connecting approach, 258, 259–260
Consolidated Criteria for Reporting Qualitative Studies (COREQ), 223, 250, 251
Constant comparative analysis, 34, 177
Constructionism, 7–8
Constructivism:
defined, 8
interpretation stage, 184
qualitative approach influence, 45–47
rigorous research approach, 221–222
socially responsible research, 9
Constructivist grounded theory, 46–47
Content analysis, 152–153
Convenience sampling, 68
Conversation analysis (CA), 39–40, 156, 157, 158 (box), 241
Corbin, Juliet, 33–34
Counter narratives, 194
Credibility, 210
Criterion sampling, 68
Critical Appraisal Skills Program (CASP), 202
Critical case sampling, 68
Critical ethnography, 46
Critical race theory, 193
Critical theories:
defined, 11, 12
interpretation stage, 193–194
Cross-national research, 92–93
Cross-sectional design, 64–65
Cross-site databases, 145

430
Cross-validation, 177
Crystallization, 216
Cultural relativism, 32

Data analysis:
action research (AR), 159–160
analytic approaches, 152–163
arts-and-crafts approach, 141
audit trail, 150
auto-coding, 143
case study analysis, 155–156
case summaries, 151
coding, 163–179
community-based participatory research (CBPR), 159–160
conceptualizing, 151–152
content analysis, 152–153
cross-site databases, 145
data management skills, 142–145
ethnographic research, 153–154
Internet resources, 182
interview transcription, 145–149
interview translation, 148–149
latent analysis, 150
longitudinal research, 160–161
manifest analysis, 150
meaningful data, 149–152
memo writing, 149–150, 166,167–169, 176–177
meta-methods, 150
multiple-user databases, 144–145
narrative approach (NA), 156–158
non-English language research, 148–149
phenomenological analysis (PA), 159
software for, 142–144
study exercises, 180
team-based research, 144
template approach, 151, 165
thematic analysis, 152
theorizing, 151–152
visual data, 161–163
Data collection:

431
 archival materials, 133–134
documents, 133–134
interviews, 108–133
observation, 99–108
saturation point, 134–135
study exercises, 136–137
Data management:
cross-site databases, 145
in data analysis, 142–145
in qualitative methods, 18–19
key word in context (KWIC), 143
multiple-user databases, 144–145
software for, 142–144
team-based research, 144
Data management log, 175–176
Data triangulation, 215
Debriefing, 80, 94, 120, 129, 148, 149, 177, see also Peer Debriefing
Debunking 243
Deception in disclosure, 79–81
DeDoose, 142
Demographic considerations:
designing research, 76
interviews, 126–127
Demystification, 243
Denzin, N. K., 5 (box), 7, 9
Derrida, Jacques, 6
Descriptive phenomenology, 41
Designing research:
conceptual framework, 60
demographic considerations, 76
design process, 62–66
ethical considerations, 57
flexibility, 58, 62
gatekeeper permission, 71, 72–74
guiding questions, 63–64, 65–66
hypotheses, 62
informed consent, 73
interviews, 72
iterative designs, 58, 62
literature review, 59–60

432
 memorandum of understanding (MOU), 74
mixed methods approach, 67, 258–263
non-English language research, 66 (box)
observation, 73–74
participant recruitment, 70–71
participant retention, 71
pilot study, 66
qualitative research approach, 61, 70, 72
rapport process, 74
research argument, 58–59
research questions, 58, 60–62
sample size, 70
sampling strategies, 67–70
self-disclosure, 75–76
self-identity, 76–77
self-presentation, 74–75
study exercises, 77–78
study questions, 58
time element, 64–65
topic selection, 58–59
Deviant case sampling, 68
Dewey, John, 8–9
Diagrams, 200
Disconfirming evidence, 198
Disconfirming sampling, 69
Discourse analysis (DA), 39, 40,156, 157, 194
Discovery of Grounded Theory,The (Glaser & Strauss), 3
Discrepant evidence, 198
Dissertation Abstracts, 203
Dissertations, 237–238
Documents:
data collection procedures, 133–134
study exercises, 136–137
Double hermeneutics, 41–42

Ecological perspective, 22
Ecological theory, 189
Ecological validity, 210, 226
Elite interview, 117–118
E-mail interview, 130

433
Emancipatory approach, 189
Embase, 203–204 (box)
Embedding approach, 258, 260–263
Emic (insider) perspective, 31–32, 95
Emotional stress:
of participant, 84–85, 129
of researcher, 93–94, 129
Empowerment theory, 22
Epiphanies, 245
Epistemology:
mixed methods research, 256
qualitative research differences, 6–8
Epoche approach, 41, 159
Ethical research:
children, 81–82
coding process, 178
Common Rule guidelines, 86
community-based participatory research (CBPR), 91
community-engaged research (CER), 90–91
confidentiality, 83–84, 88, 94
cross-national research, 92–93
deception in disclosure, 79–81
design considerations, 57
gatekeeper permission, 82, 92
goodwill payback, 85
incentive payments, 83, 85
informed consent, 73, 81–82, 87
Institutional Review Board (IRB), 86–87, 91, 92
Internet resources, 86, 91, 92, 95
moral ambiguity, 87–90
participant coercion, 82–83
participant stress, 84–85
peer debriefing, 94
privacy, 83–84
researcher safety, 94
researcher stress, 93–94
research study, 89–90 (box)
social responsibility, 95
study exercises, 96
vulnerable populations, 81–82

434
Ethnocentrism, 32
Ethnographic research:
analytic ethnography, 13
auto-ethnography, 46, 47
characteristics of, 31–33
critical ethnography, 46
cultural relativism, 32
data analysis, 153–154
emic (insider) perspective, 31–32, 95
ethics, 89–90 (box)
ethnocentrism, 32
etic (outsider) perspective, 31–32, 95
feminist ethnography, 46
field notes, 32
historical development, 3–4
holistic perspective, 32
institutional ethnography, 46
interpretation stage, 185, 199
key informants, 102
modification influences, 46, 48
participant observation, 100
performance ethnography, 46, 47
rapid ethnographic assessment (REA), 48
reports, 241
research questions, 61
research study, 32–33 (box), 38 (box)
shadowing interviews, 89
thick description, 37, 150, 241–242
Etic (outsider) perspective, 31–32, 95
European Social Work Research Association (ESWRA), 4–5 (box)
Evaluating qualitative research:
big-tent approach, 209, 225–227
checklists, 223
confirmability, 210
consensual standards, 210–211
credibility, 210
ecological validity, 210, 226
evaluative criteria, 210, 223–225
evaluative standards, 209, 210
evidentiary adequacy, 210, 226

435
 generalizability, 212
guidelines, 223
interpretive approach, 209
quasi-prescriptive approach, 209
rigor, 209, 210–212, 214–222
study exercises, 228
transferability, 210, 212–213
trustworthiness, 210, 212, 213–214
Evaluation research, 51–52
Evidence-based practice (EBP), 9–10
Evidentiary adequacy, 35, 72, 134, 210, 226
Expert interview, 117–118
Extended case method, 13
Extreme case sampling, 68

Feminist ethnography, 46
Feminist theory, 193
Field interview, 111
Field notes:
defined, 32
guidelines for, 104–105
in observation, 102–105
questions regarding, 103
research study, 103–104 (box)
Flexibility:
designing research, 58, 62
researcher skill, 18
Focused coding, 199
Focus group interview, 111–114
Follow-up interview, 72
Foucault, M., 6, 193
Funding. See Proposals for funding

Gatekeepers:
designing research, 71, 72–74
ethical research, 82, 92
Generalizability, 212
Geographic Information Systems (GIS), 107–108
Gilgun, Jane, 5 (box)
Glaser, B., 3
Go-along interviews, 100–101

436
Goffman, Alice, 89–90 (box)
Goffman, Erving, 4, 74, 244
Google Scholar, 203
Grand theories:
defined, 11
historical development, 3–4, 33
versus analytic induction, 12
Grounded theory (GT):
axial coding, 199–200
categories, 34
characteristics of, 33–35
coding process, 34, 163, 177, 199–200
constant comparative analysis, 34, 177
constructivist grounded theory, 46–47
focused coding, 199
historical development, 3, 33–34
inductive coding, 34
interpretation stage, 186, 188, 189, 197–198, 199–200, 201–202
memo writing, 34
mixed methods approach, 265 (box)
modification influences, 46–47
objectivist grounded theory, 46
of service engagement, 189
open coding, 34
reports, 240, 241
research questions, 61
research study, 35 (box)
rigorous research, 221
sample size, 70
saturation point, 34
selective coding, 199–200
sensitizing concepts, 13, 34, 165, 188
theoretical framework, 11
theoretical sampling, 197–198
value appraisal criteria, 189
Guba, E. G., 7, 256

Handbook of Qualitative Methods (Denzin & Lincoln), 7, 9

Horizontalization, 159
Hermeneutic phenomenology, 41–42

437
Holistic perspective, 32
Homogeneous sampling, 68
Horizontalization, 159
Hybrid methods approach:
characteristics of, 49
coding process, 178–179
HyperRESEARCH, 142
Hypotheses, 62

Idiographic description, 3–4

Implementation Science, 10, 269–270
Implementation science:
characteristics of, 10–11, 52
mixed methods approach, 269–270
Incentive payments, 83, 85
Individual interview, 114–115
Inductive coding, 34
Inductive reasoning, 12
Inductive theory, 12
Informal field interview, 111
Informed consent:
for ethical research, 73, 81–82, 87
in research design, 73
Instagram, 105–106
Institutional ethnography, 46
Institutional Review Board (IRB), 86–87, 91, 92
Instrumental case study, 36, 37, 155, 199
Intensity sampling, 68
Inter-coder agreement, 164
Interdisciplinary triangulation, 215
International Compilation of Human Research Standards (ICHRS), 92
International Congress of Qualitative Inquiry (ICQI), 5 (box)
International Institute for Qualitative Methodology, 279
International Journal of Qualitative Methods, 130
Internet interview, 130–132
Internet publishing, 249
Interpretation stage:
action research (AR), 184, 199
analogies, 194–195
case study analysis, 199

438
 change-oriented research, 185
coding process, 199–200
conceptualizing, 190, 191f
constructivist approach, 184
critical perspective, 193–194
disconfirming evidence, 198
discrepant evidence, 198
ethnographic research, 185, 199
framing devices, 185–186, 195
grounded theory (GT), 186, 188, 189, 197–198, 199–200, 201–202
Internet resources, 202, 204
metaphors, 194–195
meta-synthesis approach, 202–205
methodological transparency, 206
narrative approach (NA), 194, 199
negative case analysis, 197–198
nonchange-oriented research, 185
phenomenological analysis (PA), 199
qualitative research approach, 198–202
researcher role, 184–185
research study, 187–188 (box), 195–197 (box), 201 (box), 204–205 (box)
rhetorical devices, 194–197
salience, 183, 190, 192, 195
saturation point, 187–188
software for, 186
study exercises, 207
systematic review, 202
thematic analysis, 200–201, 202
theorizing, 188–190, 195–197 (box)
Interpretive phenomenological analysis (IPA), 41–42
Interview feedback form (IFF), 124–126
Interview guide, 118–122
Interviews:
audio recording, 132
authenticity of, 108–110
children, 116–117
computer-generated interview, 130–132
Critiques of, 108–110
data collection procedures, 108–133
demographic considerations, 126–127

439
 elicitation techniques, 120–121
elite interview, 117–118
e-mail interview, 130
expert interview, 117–118
focus group interview, 111–114
guidelines for, 114 (box), 122
individual interview, 114–115
informal field interview, 111
Internet interview, 130–132
interview feedback form (IFF), 124–126
interview guide development, 118–122
key informants, 117
limitations of, 108–110
linked interview, 116
minimally structured, 115
multiple interviews, 121–122
non-English language, 129–130, 148–149
participant stress, 129
pilot test, 122
probes, 123–124
realism of, 108–110
researcher errors, 127–128
researcher stress, 129
research study, 113 (box), 123 (box), 132 (box)
shadowing interviews, 89, 100–101
study exercises, 136
telephone interview, 130
transcription of, 145–149
translation of, 148–149
video recording, 133
vulnerable populations, 116–117
Intrinsic case study, 36, 37, 155, 199
In vivo codes, 170, 190, 191t, 195
Irony, use of 243–244
Iterative designs, 58, 62, 183–184

Journal publications, 246–249

Journal of Mixed Methods Research, 255, 271
Journal of Phenomenological Psychology, 241
Journal of Video Ethnography, 105

440
Journal resources, 4, 5 (box), 10, 26–27, 44–45 (box), 46, 105, 130, 241, 249–250, 255, 269–270, 271

Key informants, 102

Key word in context (KWIC), 143

Labov, William, 40,156

Liebow, E., 4, 194
Life course studies, 160–161
Life histories, 160–161
Lincoln, Y. S., 7, 9, 256
Linked interview, 116
Literature review, 59–60
Longitudinal research:
case examples, 22–24
data analysis, 160–161
life course studies, 160–161
life histories, 160–161
time element, 64–65

Mapping technique, 107–108

Master narratives, 194
Maximum variation sampling, 68
Mead, George Herbert, 8–9
Mead, Margaret, 105
Meaning units, 164
Member checking, 217f, 219, 221–222
Memorandum of understanding (MOU), 74
Memo writing:
code notes, 176
coding process, 176–177
data analysis, 149–150, 166, 167–169, 176–177
grounded theory (GT), 34
operational notes, 176
theory notes, 176
Merging approach, 258, 260–262
Meta-methods for analysis, 150
Metaphors:
interpretation stage, 194–195
reports, 245
Meta-synthesis approach, 202–205
Methodological transparency, 206

441
Methodological triangulation, 215
Mixed methods research:
action research (AR), 269, 270–271
as qualitative approach, 47, 48–50
as third paradigm, 255, 256
challenges of, 258–259, 273–274
coding process, 178–179
community-based participatory research (CBPR), 270–271
complicated design approach, 262–263, 269–270
concurrent design, 258, 260–262, 264
connecting approach, 258, 259–260
design considerations, 67, 258–263
development of, 257
embedding approach, 258, 260–263
epistemology question, 256
for program evaluation, 275–276
grounded theory (GT), 265 (box)
implementation science, 269–270
integration options, 257–258
merging approach, 258, 260–262
quantifying qualitative data, 265 (box), 266–269
rationale for, 257
reports, 271–272
research challenges, 273–275
research process, 263–272
research study, 260–261 (box), 262–263 (box), 265 (box), 268–269 (box), 274–275 (box)
sequential design, 258, 259–260, 264, 271
standardized measures, 266
structural design decisions, 263–266
study exercises, 277
Mixing qualitative approaches, 67, 178–179
Moral ambiguity, 87–90
Morrow, Susan, 35
Multiple interviews, 121–122
Multiple-user databases, 144–145

Narrative approach (NA):

characteristics of, 38–40
conversation analysis (CA), 39–40, 156, 157, 158 (box), 241
counter narratives, 194

442
 data analysis, 156–158
discourse analysis (DA), 39, 40, 156, 157
emplotment, 157
interpretation stage, 194, 199
master narratives, 194
mixed methods research, 47
modification influences, 46
notation system transcription, 158 (box)
reports, 241
research questions, 61
rigorous research, 221
zoom-in/zoom-out technique, 199, 241
National Association of Social Workers Code of Ethics, 95
National Institutes of Health (NIH):
Federal Certificate of Confidentiality (CoC), 84
Naturalistic Inquiry (Lincoln & Guba), 7
Negative case analysis:
interpretation stage, 197–198
rigorous research, 217f, 220, 221
New York Recovery Study (NYRS):
case study analysis, 155
case summaries, 151
coding process, 166, 167–168t
confidentiality, 83
data management, 143–144
description of, 22–24
field notes, 103–104 (box)
gatekeepers, 72
geo-coding, 108
goodwill paybacks, 85
longitudinal design, 160–161
mixed methods research, 267, 268–269 (box)
moral ambiguity, 87, 88
participant stress, 84–85
peer debriefing and support (PDS), 217–218
photo-elicitation interview (PEI), 106–107 (box), 108, 161–163
reports, 235, 236–237, 242
researcher safety, 94
researcher self-presentation, 74, 75
researcher stress, 94, 129

443
 shadow interviews, 101
New York Services Study (NYSS):
codebook, 171–175
description of, 22
gatekeepers, 72
interpretation stage, 189, 190, 191t, 192, 193–197
mixed methods research, 267
peer debriefing and support (PDS), 217–218
reports, 236
researcher self-presentation, 74
Non-English language research:
designing research, 66 (box)
interviews, 129–130, 148–149
translation accuracy, 148–149
Notation system transcription, 158 (box)
Numbers, use of, 234–235
NVIVO, 142

Objectivism, 46
Objectivist grounded theory, 46
Observation:
approach to, 101–105
data collection procedures, 99–108
designing research, 73–74
field notes, 102–105
go-along interviews, 100–101
key informants, 102
mapping technique, 107–108
participant observation, 100
participatory mapping, 108
photo-elicitation interview (PEI), 105, 106–107, 108, 161–163
photography, 105–108
photovoice (PV) techniques, 105–106
reactivity, 100
research study, 103–104 (box), 106–107 (box)
shadow interviews, 100–101
study exercises, 136
video recording, 105–108
Observer triangulation, 215
On the Run (Goffman), 89–90 (box)

444
Open coding, 34, 164–166, 167–168t
Open systems theories, 2, 11, 13, 189
Operationalism, 5
Operational notes, 176
OVID Medline, 203–204 (box)

Paradigm wars, 6–8

Park, Robert, 3
Participant observation, 100
Participants:
coercion of, 82–83
emotional stress, 84–85, 129
recruitment of, 70–71
respondent bias, 213–214, 217f
retention of, 71
Participatory action research (PAR):
characteristics of, 42–43, 44
modification influences, 47
research questions, 61
Participatory mapping, 108
Peer debriefing and support (PDS), 94, 216–218, 221–222
Peer review, 245–246, 247
Peirce, Charles, 8, 13
Performance ethnography, 46, 47
Phenomenological analysis (PA):
characteristics of, 41–42
data analysis, 159
descriptive phenomenology, 41
double hermeneutics, 41–42
epoche approach, 41, 159
hermeneutic phenomenology, 41–42
interpretation stage, 199
interpretive phenomenological analysis (IPA), 41–42
modification influences, 46
reports, 240–241
research questions, 61
rigorous research, 221
sample size, 70
transcendental phenomenology, 41
Photo-elicitation interview (PEI), 105, 106–107, 108, 161–163

445
Photography, 83, 105–108
Photovoice (PV) techniques, 105–106
Pilot study, 66, 285
Pragmatism:
characteristics of, 8–9
qualitative approach influence, 47
Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA), 202, 204
Privacy, 83–84, 146
Probes, 118, 123–124
Program evaluation:
mixed methods approach, 275–276
qualitative approach, 51–52
Prolonged engagement, 214–215, 217f, 221, 222
Proposals for funding:
budget justification, 284
budget resources, 283–284
checklist for, 284–285
criteria for, 282
federal agencies, 279, 281, 288
foundation support, 280
Internet resources, 279–280
novice challenges, 281–282
organizational interests, 280–281
research contractors, 281
reviewer questions, 282–283
time allocation, 283
workshops, 279
Prospective design, 64
Prospective Register of Systematic Reviews (PROSPERO), 203–204 (box)
PsychINFO, 203–204 (box)
Publishing, 245–249
PubMed, 203
Purposive sampling:
qualitative approach influence, 47
types of, 68–69

QDA Miner, 142

QDAP, 142
Qualitative approach:
action research (AR), 42–43, 47

446
 case study analysis, 35–38
community-based participatory research (CBPR), 42–45, 47
community-engaged research (CER), 43, 44
constructivism influence, 45–47
designing research, 61, 70, 72
ethnographic research, 31–33, 46, 48
grounded theory (GT), 33–35, 46–47
hybrid research, 49
implementation science, 52
in program evaluation, 51–52
interpretation stage, 198–202
mixing qualitative approaches 47, 48–50
modification influences, 45–48
narrative approach (NA), 38–40, 46, 47
objectivism influence, 46
participatory action research (PAR), 42–43, 44, 47
phenomenological analysis (PA), 41–42, 46
postmodernism influence, 45–46, 47
pragmatism influence, 47
reports, 240–241
research study, 32–33 (box), 35 (box), 38 (box), 44–45 (box)
rigorous research, 50–51, 221–222
study exercises, 53
trustworthy research, 50
Qualitative comparative analysis (QCA), 12
Qualitative Data Analysis (QDA) software:
for data interpretation, 186
for data management, 142–144
Qualitative Health Research, 249
Qualitative Inquiry, 46
Qualitative methods:
abduction, 13, 188–189
analytic induction, 12–13, 188
conceptual framework, 11, 13–14
constructionism, 7–8
constructivism, 8, 9, 45–47, 184, 221–222
data management skills, 18–19, 142–145, 175–176
developmental history, 3–11
epistemological differences, 6–8
evidence-based practice (EBP), 9–10

447
 familiar-unfamiliar topics, 20–21
flexibility skill, 18
formative years, 3–4
implementation science, 10–11
Internet resources, 5 (box), 9, 27
journal resources, 4, 5 (box), 10, 26–27
longitudinal research studies, 22–24
open systems theories, 2, 11, 13, 189
operationalism, 5
pragmatism, 8–9
quantification contrast, 1–3
quantification development, 5–6
realism, 6
reflexivity skill, 18
reasons for doing, 16–18
researcher skills, 18–19
research process, 15f
research scenarios, 16–18
socially responsible research, 21–22
social work research development, 4–5 (box)
study exercises, 24–25
theoretical framework, 11–16
theoretical questions, 14–16
Qualitative Psychology, 4
Qualitative Social Work, 5 (box), 249–250
Quantitative methods:
historical development, 5–6
qualitative contrast, 1–3
reporting numerical findings, 234–235
Queer theory, 193
QUIRKOS, 142

Rapid ethnographic assessment (REA), 48

Rapport process, 74
RATS guidelines, 250, 251
Reactivity:
in observation, 100
trustworthiness threat, 213, 217f
Realism, 6
Reflective pragmatism, 9

448
Reflexivity 18, 23, 46, 47, 50, 108, 176, 184, 185, 193, 213, 216, 224, 227, 234, 244, 250
Reports:
abstract section, 239
audience identification, 232–233
authorship, 231, 235–238
background/theoretical context section, 239
case study analysis, 241
case vignettes, 242
checklist for, 249–251
coauthorship, 231, 235–238
conclusions/recommendations section, 240
debunking device, 243
demystification device, 243
description-interpretation balance, 241–242
dissertations, 237–238
epiphanies, 245
ethnographic research, 241
findings section, 240
format organization, 239–240
grounded theory (GT), 240, 241
irony device, 243–244
journal publications, 246–249
metaphors, 245
methods section, 239–240
mixed methods research, 271–272
narrative approach (NA), 241
numerical findings, 234–235
online publishing, 249
peer review, 245–246, 247
phenomenological analysis (PA), 240–241
preparation decisions, 232–238
professional jargon, 244–245
publication challenges, 245–249
qualitative research approach, 240–241
researcher role, 233–234
rhetorical devices, 243–244
rhythmic quality, 244
study exercises, 252
title section, 239
turning points, 245

449
 writers, 231
writing style, 244–245
writing suggestions, 232
Researchers:
bias, 213, 217f
emotional stress, 93–94, 129
flexibility, 18
reflexivity, 18
research skills, 18–19
role of, 184–185, 233–234
safety, 94
self-disclosure, 75–76
self-identity, 76–77
self-presentation, 74–75
Research Portfolio Online Reporting Tools (RePORT), 280–281
Research questions, 58, 60–62
Respondent bias, 213–214, 217f
Respondent-driven sampling (RDS), 69
Retrospective design, 64
Retrospective interpretation, 177
Rhetorical devices:
interpretation stage, 194–197
reports, 243–244
Rigor in qualitative research:
action research (AR), 221, 222
audit trail, 217f, 220, 221
case study analysis, 221
community-based participatory research (CBPR), 222
constructivist approach, 221–222
defined, 211–212
differing perceptions of, 50–51
grounded theory (GT), 221
member checking, 217f, 219, 221–222
narrative approach (NA), 221
negative case analysis, 217f, 220, 221
peer debriefing and support (PDS), 216–218, 221–222
phenomenological analysis (PA), 221
prolonged engagement, 214–215, 217f, 221, 222
qualitative research approach, 221–222
strategies for, 50–51, 214–222

450
 triangulation of data, 215–216, 217f, 221
verisimilitude, 50

Safety of researcher, 94
Salience, 183, 190, 192, 195
Sample size, 70
Sampling strategies:
analysis-driven techniques, 69
designing research, 67–70
purposive sampling techniques, 67–69
Saturation:
data collection, 64 134–135
grounded theory (GT), 34
interpretation stage, 187–188
sample size, 70
Schemas, 190, 200
Selective coding, 199–200
Self-disclosure of researcher, 75–76
Self-identity of researcher, 76–77
Self-presentation of researcher, 74–75
Sensitizing concepts, 13, 34, 165, 188
Sequential design, 258, 259–260, 264, 271
Serendipitous findings, 274–275
Shadowing interviews, 89, 100–101
Situational analysis, 13
Snowball sampling, 68–69
Social Ecology Theory, 11, 13
Socially responsible research:
constructivism, 9
ethics, 95
qualitative methods, 21–22
social justice values, 21–22
Social work and qualitative methods, 4–5
Society for Social Work and Research (SSWR), 4–5 (box), 279
Standards for Reporting Qualitative Research (SRQR), 250
Statistics, 5–6
Strategies for rigor:
applied, 221–222
definition, 209
types of, 214–220

451
Strauss, A., 3
Study questions, 58
Substance Abuse and Mental Health Services Administration (SAMHSA), 281
Survey research, 5–6
Symbolic interactionism, 189
Symbolic Interaction Theory, 11, 13
Systematic review, 202

Talley’s Corner (Liebow), 4, 194

Team-based research, 144
Telephone interview, 130
Template approach, 49, 151, 159, 165, 165, 179, 200, 263, 272
Thematic analysis:
data analysis approach, 152
interpretation stage, 200–201, 202
qualitative approach influence, 47
Theoretical codes, 170
Theoretical framework:
characteristics of, 11–16
questions for, 14–16
Theoretical questions:
“how” question, 14, 15–16
“what” question, 14–15
“when” question, 15
Theoretical sampling, 69, 197–198
Theorizing:
data analysis, 151–152
interpretation stage, 188–190, 195–197 (box)
Theory notes, 176
Theory triangulation, 215
Thick description, 37, 150, 189, 211, 241–242
Time element:
proposals for funding, 283
research design, 64–65
Topic selection:
familiar-unfamiliar topics, 20–21
in research design, 58–59
Transcendental phenomenology, 41
Transcribing interviews:
for data analysis, 145–149

452
 guidelines, 146–147
non-English language research, 148–149
transcriber error, 147–148 (box)
Transferability, 210, 212–213
Translating interviews, 148–149
Triangulation of data, 215–216, 217f, 221
Trustworthiness:
defined, 50, 210, 212
researcher bias, 213, 217f
respondent bias, 213–214, 217f
Threats to, 213–214
Turning points, 245
Typical case sampling, 68
Typologies, 37

U.S. Office for Human Research Protections (OHRP), 86, 92

Values in qualitative research, 21–22

Verisimilitude, 50
Verstehen description, 189
Video recording:
confidentiality, 83–84
for interviews, 133
in observation, 105–108
Visual anthropology, 105
Visual data analysis, 161–163
Visual sociology, 105
Vulnerable populations:
ethical research, 81–82
interviews, 116–117

Web of Science, 203

Wolcott, Harry, 4

Zoom-in/zoom-out technique, 199, 241

453
About the Author

Deborah K. Padgett
is widely known for her expertise in qualitative and mixed methods including this third edition and a
companion text Qualitative and Mixed Methods in Public Health (2012). She is first author of Housing
First: Ending Homelessness, Transforming Systems and Changing Lives (2016) and has published
extensively on mental health needs and service use of homeless adults and other underserved
populations. Dr. Padgett was principal investigator on two NIMH-funded R01 grants, both qualitative
studies of engagement and recovery among formerly homeless adults in supportive housing in New York
City. She served as president of the Society for Social Work and Research (SSWR), culminating in the
establishment of the SSWR “Deborah K. Padgett Early Career Award” in recognition of her
contributions. She is a member of the American Academy of Social Work and Social Welfare
(AASWSW). In 2013, Dr. Padgett received New York University’s Distinguished Teaching Award.

454