Sie sind auf Seite 1von 26

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/38415151

Types of Case Study Work: A Conceptual


Framework for Case-Based Research

Article in Journal of Humanistic Psychology July 1998


Impact Factor: 0.55 DOI: 10.1177/00221678980383003 Source: OAI

CITATIONS READS

68 1,594

1 author:

David Edwards
Rhodes University
84 PUBLICATIONS 700 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: David Edwards
letting you access and read them immediately. Retrieved on: 30 June 2016
Citation:
Edwards, D. J. A. (1998). Types of case study work: A conceptual framework for case-based research.
Journal of Humanistic Psychology, 38(3), 36-70.

A pdf of this article as published is available online at http://jhp.sagepub.com/cgi/reprint/38/3/36

Types of case study work: A conceptual framework for case-based research1

David J. A. Edwards2
Rhodes University, Grahamstown, South Africa

Abstract
This paper describes a conceptual framework for understanding the phases of case-
based research. Case-based strategies in research are widely used in case study
methodology as well as in a number of qualitative methodologies including
grounded theory development, phenomenological research method and
psychotherapy process research. The epistemological principles upon which case-
based research is based are fundamentally different those which inform group-
based research using quantitative multivariate statistics. The case-based research
process is divided into three general phases: descriptive, theoretical-heuristic, and
theory-testing. Each of these phases is subdivided into two categories. The aims
and epistemological principles related to each phase are discussed and illustrated
with examples from contemporary research. The principles for establishing validity
in case-base research are also reviewed and, in order to assist research students
and supervisors, some common pitfalls are noted. The paper shows that rigorous
work with single cases or series of cases using case-based principles and
methodology is indeed, as Bromley (1986) claimed, "the bedrock of scientific
investigation".

The purpose of this paper is to present a conceptual framework for guiding the process of
case-based research. In this type of research, one or more cases of a phenomenon of
interest are systematically examined with a view to achieving an understanding and
developing or extending a theoretical framework. A case study is a case-based research
project which examines a single case, usually in considerable depth. In a multiple case
study, a series of cases is examined. However the case-based approach is not limited to
research formally bearing the title case study. Many qualitative research approaches use
case-based strategies. These can be contrasted with the strategies of group-based
quantitative research, where conclusions are drawn by means of statistical analysis of
groups of cases whose scores are combined in descriptive statistics such as the mean or
variance.

Quantitative multivariate methods have the advantage of allowing researchers to measure


and control variables, but have the disadvantage that the resulting theory often fails to take
account of the unique characteristics of individual cases. In clinical psychology, it has long
been recognised that the group-based research process failed to contribute the kind of
knowledge base that is of direct application to much of clinical practice. In 1981, a special

1
issue of the Journal of Consulting and Clinical Psychology addressed this concern (Barlow,
1981; Hayes, 1981; Kiesler, 1981; Strupp, 1981). More recently, several papers on the
practice of psychotherapy have emphasised the gulf between the problems and concerns of
practitioners and the findings of group-centred research (Mahrer, 1988; Polkinghorne,
1986; Safran, Greenberg & Rice, 1988).

Thirty years ago, M. B. Shapiro (1966, p. 19) was a rather lone voice when he argued for the
importance of individual-centred research as a complement to group-centred research.
At that time I was receiving my own training in psychological research methods which was
entirely in the form of multivariate group-centred strategies. Like many students then, I
felt a deep discomfort with the way these methods distanced the researcher from the
psychological phenomena under study and resulted in conclusions which failed to do justice
to the full complexity and dignity of what it is to be human. Later, I trained in the field of
cognitive therapy and began to reflect on how Beck and his colleagues (Beck, Rush, Shaw
and Emery, 1979; Beck & Emery, 1984) had derived the elaborate set of constructs and
principles upon which this clinical practice was based. The answer was easy to find. It was
the careful observation, description and comparative discussion of individual cases that was
the foundation of the development of clinical knowledge (Edwards, 1996a). I realized that
this process was, in fact, the foundation of knowledge in the human and social sciences in
general.

Case studies have played a pivotal role in the evolution of humanistic and transpersonal
clinical methods3. Robert Whites (1972) Lives in Progress, now out of print, is a classic
collection of pioneering and compelling psychological case studies in the personological
tradition of Henry Murray and Gordon Allport. Carl Rogers published the first case study
promulgating a person-centered approach to therapy in contrast to the prevailing
behavioural and psychoanalytic alternatives. His case of Herbert Bryan (Rogers, 1942) was
particularly notable for incorporating an actual transcript from the session to illustrate the
unique aspects of this approach. According to Bugental (personal communication to David
Lukoff), a leading proponent of existential psychology, Rogers inclusion of a verbatim
transcript caused quite a stir. Prior to that, published case studies had never included the
actual exchanges between therapist and client. Rollo Mays (1972) case of Mercedes, a
black woman of low socio-economic status, established the applicability of a humanistic-
existential approach to a broader range of clientele than that addressed by psychoanalysis.
Ever since Kraepelins work, and even before, the process of developing a system of
diagnostic categories has drawn extensively on case studies. The impact that case studies
continue to have in this area is illustrated by the recent development and acceptance of
the new diagnostic category of religious or spiritual problem (American Psychiatric
Association, 1994). A published case study of a spiritual emergency (Lukoff, 1985; Lukoff &
Everest, 1985) played a pivotal role in the evolution and adoption of this category (Lukoff,
Lu & Turner, 1998).

Historically, case-based research has always had a place in psychology (Bolgar, 1965) and
the social science literature of the 1930's contains many expositions of its principles.
However, recognition of its value was increasingly lost as enthusiasm grew for the
development of quantitative methodologies. In most social sciences, it came to be

2
completely discredited, and it survived mainly in anthropology and psychoanalysis
(Mitchell, 1983). Over the past three decades, of course, there have been many
independent moves to correct the balance and to legitimize case-based research models.
One aspect of this has been the recognition that many of the methods of clinical practice
constitute a systematic, if informal, research process. Several writers have pointed out that
clinical psychological assessment is a form of single case research (e.g. Kiesler, 1981;
Kratochwill et al, 1984; Shapiro, 1966). The clinician is usually motivated to answer a
specific question, analogous to a research question, and gathers data in a systematic
manner that is designed to provide an answer to the question. This means that the
principles by which conclusions are drawn from clinical assessments are of relevance when
examining how conclusions are drawn from cases studies in general. A second important
aspect has been the development of single case experimental designs (Hayes, 1981; Barker,
Pistrang & Elliott, 1994; Kratochwill, Mott & Dodson, 1984). Another has been the
emergence of case-based qualitative research methodologies (Miles & Huberman, 1994;
Taylor & Bogdan, 1984), and, in particular, phenomenological research method (Giorgi,
1985), discovery-oriented psychotherapy research (Mahrer, 1988), grounded theory
building (Glaser & Strauss, 1967), co-operative experiential enquiry or new paradigm
research (Reason, 1988), psychotherapy process research (Rice & Greenberg, 1984) and
case study research (Bromley, 1986; Edwards, 1990, 1996a; Stake, 1994, 1995; Yin, 1984).

By and large, case-based approaches share six assumptions:


1. The ultimate task of science is not to map human behaviour and experience on to a
vast matrix of variables from which predictions can be derived. Factor analytic
models, for example, are based on the untested assumption that human behaviour
and experience can be modelled as a set of discrete variables which are mutually
determined in a predictable way (Polkinghorne, 1986). As a result, the
mathematical models on which factor analyses are based bear only a limited
relationship to real psychological and interpersonal processes.

2. It follows, therefore, that quantification, although at times it may have a place, is


not an end in itself. Individuals are not merely sources of scores from which can be
estimated the means and variances of population distributions or the slopes of
regression lines.

3. The quality of our science depends on the quality of our data. The limited value of
factor analytic personality models is partly due to the fact that the data base is quite
impoverished compared to the in depth case material which forms the foundation
of case-based theories. Josselson and Lieblich (1993, p. ix) express it thus:
Listening to people talk in their own terms about what had
been significant in their lives seemed to us far more valuable
than studying preconceived psychometric scales or contrived
experiments.

4. Participants are encountered respectfully as persons and treated as individuals


whose accounts of themselves call for exploration and understanding. In some
models they are invited to participate as collaborative partners or co-researchers

3
(Aldridge, 1994; Reason, 1988).

5. Data are contextualized. This means that a sufficient range of qualitative data are
collected to enable meaningful relationships to be examined within a single case.

6. Research needs to be case-based and case-centred. This means that the material of
each case is taken seriously and used as a basis for theory development. The
resultant theory can in turn be brought profitably to bear in understanding and
researching new cases.

Although case studies can utilise any form of relevant data, in the present paper the focus is
largely on phenomenological case studies, that is to say case studies in which the central
research question focuses on participants' experience. This is, of course, the focus of much
psychotherapy research, from which I will draw many illustrations of the principles. Not all
qualitative data provide access to participants' experience. For example, statements of
attitude and belief, especially when held rigidly and in a stereotyped manner, actually
obscure the nature of the underlying experience. By contrast, observable behaviour can be
treated as part of phenomenological data if it is viewed not as a mechanistic response, but
as a meaningful incarnation of an individual's lived reality (as explicated, for example, by
Merleau-Ponty, 1967).

Implicit in all the research approaches mentioned are common underlying perspectives on
the research process itself. It is the purpose of this paper to present an integrative
framework which articulates these perspectives as a series of recognizable stages governed
by identifiable epistemological principles. When researchers do actual case studies, they
must select specific strategies and techniques for data collection, analysis and
interpretation. Several examples of such specific techniques will be referred to by way of
illustration. However, this paper is not a manual of how to execute specific research tasks.
It is, rather, a metatheoretical conceptualisation designed to facilitate the selection,
implementation and evaluation of specific strategies and techniques.

The focus of the paper is not specifically on qualitative research, although most of the
research to be discussed is qualitative in nature. Rather it is on the centrality of the study
of individual cases to the research enterprise. The general aim of this paper is to restate
some tested principles of a long tradition of case-based research in psychology. In the
process, a set of concepts will be set out which will help direct researchers to the kind of
issues that need to be attended to as they move through the stages of a case-based
research project. This in turn will facilitate the identification of some of the methodological
flaws which must be avoided if case-based research is to yield meaningful and valid
conclusions.

Erik Eriksons biographies of Ghandi (Erikson, 1969) and Luther (Erikson, 1993) are examples
of psychobiography. This is a particular kind of case-based research which embraces the
whole courseof an individuals whole life, or a large part of it (Denzin, 1989; Rosenthal,
1993; Smith, 1994). Many of the principles discussed here apply to biographical writing.
However, this paper will largely deal with more focused kinds of case-based research, many

4
of which are clinical case studies. In clinical case studies, it is often possible to work more
with more theoretical rigor and to rely less on poorly supported interpretative inferences.
This is because they not only have a more narrow focus, but also because the researcher or
clinician is usually in a position to gather further information in order to clarify details and
check on hypotheses and interpretations. By contrast, biography is often written post hoc
from archival data sources and the kinds of information that would be required to verify
some interpretations may not be obtainable. Nevertheless there is considerable overlap
between the task of writing an extended biographical account of a persons life and writing
a focused description of a circumscribed part of it. This is clear from the fact that, a central
feature of most clinical case studies is a case history which is used to situate and offer an
understanding of critical events in the present. Another pointer to the degree of overlap is
the range of methods used in readings in the series The narrative study of lives (Josselson &
Lieblich, 1993, 1995; Lieblich & Josselson. 1994). Weilands (1995) account of the career of
a 50 year old college professor is a biographical study that focuses on the interaction
between the development of an individual career and the societal context in which it is
situated, over a period of some 30 years. By contrast, Lieblichs (1993) study of the
experience of a 21 year old woman as a new immigrant to Israel from Russia which, like
many clinical case studies situates a psychological investigation of her current experience
against the background of her previous history. An even narrower time range is involved in
Rogers, Brown & Tappans (1994) case studies of adolescent girls based on annual
interviews over a five year period.

CASE STUDY WORK AT DIFFERENT PHASES OF THE RESEARCH PROCESS


In The farther reaches of human nature, Maslow (1973, p.4), contrasts the task of
"pioneering, scouting, originating", with that of "validating, checking, verifying". Similarly,
Glaser and Strauss (1967) distinguish between discovering theory and testing theory. These
represent two fundamental phases of the research process, the context of discovery and
the context of verification (Giorgi, 1986b; Henwood & Pidgeon, 1993). Since, historically,
the context of verification has received disproportionate emphasis, there has been a loss of
respect for the context of discovery. Giorgi (1986b) has argued for the importance of
restoring descriptive science as the foundation of the process of discovery, without which
the elaborate machinery of verification is meaningless.

The research process can, in fact, be usefully conceptualised as having three major phases
(Eckstein, 1975; Giorgi, 1986a): a descriptive phase, a theoretical-heuristic or theory
development phase, and a theory-testing phase. It is not only at the descriptive phase that
case-based research is applicable. It contributes at every phase of the research process.
The phases are separately demarcated for the purpose of presentation. In practice, they
form a continuum, since description always involves some preliminary theory development
and theory development always involves some preliminary theory testing. Most case
studies include more than one phase since they include descriptive work as well as
examination of theory. I have divided each phase into two sub-classes, so that altogether
there are six categories, which I have found useful in supervising postgraduate case study
research. These are summarised in Table 1. Each category represents a different aspect of
the research process, has specific kinds of aims, and suggests specific kinds of strategies.
For this reason, I refer to each one as involving particular types of case study work. When

5
researchers fail to separate out the different kinds of case study work, their reports are
likely to lack clarity, direction or coherence. The quality of presentation and the dignity of
the method will be enhanced where researchers cultivate (and make clear to the reader) an
awareness of the particular kind of case study work they are doing in each portion of their
report.

Table 1
Categories of case study work
Exploratory-descriptive work
Descriptive work
Focused-descriptive work

Grounded theory building:


Categorization
Statements of correlational relationship
Theoretical-heuristic Definition of psychological structure
Propositions about processes
work
Propositions about causal relationships
Hermeneutic work

Testing propositions within grounded theory


Theory testing work
Metatheoretical deconstruction

Illustrations of how these principles have been applied in practice will be drawn from
several sources. Three in particular will be referred to quite frequently. The first is the
phenomenological research method developed at Duquesne University in Pittsburgh
(Edwards, 1991; Giorgi, 1985). The method is designed to furnish a systematic and rigorous
approach to the understanding of the psychological structure of human experiences. It has
been used to study diverse experiences including learning (Giorgi, 1985), anger (Stevick,
1971), guilt (Brooke, 1985a; 1985b), anxiety (Fischer, 1982), suspicion (de Koning, 1979), a
woman's premenstrual phase (Montgomery, 1982), being an immigrant (Polyzoi, 1985),
being the victim of a crime (Wertz, 1985), having one's house attacked and damaged
(Cleaver, 1988), being a therapist working with a client's projective identification (Thorpe,
1989), and achieving an insight in psychotherapy that brings a sense of freedom (Todres,
1990).

Second, examples will be taken from the steps described by Mahrer (1988) in his account of
discovery-oriented psychotherapy research. This is one example of the evolving field of

6
process research in psychotherapy (Greenberg & Pinshof, 1986; Rice & Greenberg, 1984).
Mahrer examines how one would research the question, "What is happening when a
[psychotherapy] patient erupts into spontaneous hearty laughter?" The approach includes a
five step strategy to promote the discovery phase of investigation and also provides
guidelines for rigorous theory building.

Third, examples will be taken from the Sheffield Psychotherapy Research Project (Stiles,
Elliot, Llewelyn, Firth-Cozens, D. Shapiro, & Hardy, 1990; Stiles, Morrison, Haw, Harper, D.
Shapiro, & Firth-Cozens, 1991). One aim of this project was to describe the process of
assimilation of emotionally distressing material (impulses, memories, emotions, beliefs)
during the course of psychotherapy. It gave rise to the development of the assimilation
model which describes the changes that take place as an individual moves from avoiding
and warding off the material to mastering and integrating it.

PHASE ONE: DESCRIPTIVE WORK


Careful, unbiased description is essential when investigating a new or little researched
phenomenon. This task calls for detailed observation of the phenomenon and the
identification of basic concepts and distinctions which will enable what has been observed
to be described accurately. In the first step of research in a new area, a set of valid cases of
the phenomenon must be obtained, and then carefully documented and examined to
furnish a thorough description. Safran et al. (1988, p.15) point out that this phase of
research is
a conceptually demanding, methodologically rigorous, and
labour-intensive process which should not be relegated to the
status of "pilot work" which takes place before the "real"
research begins.
Mahrer recommends that, to study a phenomenon like a therapy client bursting into
spontaneous laughter, the researcher should obtain instances of the phenomenon in the
form of video or audio tapes of therapy sessions. This may require ingenuity, the seizing of
opportunities that present themselves, or access to a comprehensive therapy tape library.
The selected cases serve as the basis for the development of a differentiated description
which opens up an understanding of something that was previously only superficially
known. Mahrer's (1988, p. 698) advises the researcher to "use terms that are close to the
data and are simple, rather than using high-level constructs wrapped in the jargon of any
therapeutic approach". This stage particularly honours Husserl's dictum "Back to the things
themselves" (Kruger, 1988, p. 28).

Exploratory-descriptive work
In exploratory-descriptive work, a relatively unknown phenomenon is examined and
investigated on its own terms in this kind of open-minded manner. Eckstein (1975) calls
this stage "configurative-idiographic". It is idiographic because the goal is not to generalise
to other cases or to develop theory. It is configurative in that the researcher strives to
achieve an organised and coherent presentation of the phenomenon.

Focused-descriptive work
In focused-descriptive work, a phenomenon that has already been described is examined in

7
more detail and researchers bring to bear specific questions about the nature of the
phenomenon. Focused descriptive work is the basis of the Duquesne phenomenological
approach. Researchers ask a carefully phrased question which directs the respondent to the
type of experience the researcher is interested in. The raw data are gathered in the form of
tape recorded interviews, or written protocols. Further questions are generally asked,
aimed at getting respondents to elaborate on particular aspects of their responses or to
speak about omitted aspects. Usually data from four to ten respondents are used for
in-depth analysis. A systematic analysis of the data into natural units of meaning is the first
step towards writing a "situated structure" of each respondent's experience. This is a data
reduction in the form of a third person synoptic summary of the individual's experience,
stripped of redundancy, coherently organised and rendered into psychological language so
that the underlying psychological dimensions of the experience are rendered explicit

Similarly, Mahrer describes how to go about developing a system for analysing and
categorising the cases that have been identified (of psychotherapy clients erupting into
laughter). He suggests that this could involve reviewing the audio and video recordings
again and again to become thoroughly familiar with them; writing detailed synopses of each
case example; using other researchers as judges to review the material; becoming
thoroughly familiar with the relevant literature so as to be alerted to what to look for;
conducting additional interviews with the participants to obtain information to supplement
what is recorded (for example their thoughts or feelings at specific times).

In case study research, the advantage of working with a single participant is that resources
are more likely to be available to obtain information from a variety of sources. A single case
study might include data in the form of physiological measures, observations of behaviour,
psychometric test results, statements of informal observers (such as family members), and
interview material (Bromley, 1986). This enables descriptive data to be contextualized, so
that specific pieces of information can be more reliably interpreted later.

Description cannot be entirely separated from theory development, and, in this phase,
aspects of a case which are investigated and described in depth are to a greater or lesser
extent theory driven. For example, Grayson, Foa and Steketee (1985) classified obsessive-
compulsive individuals on the basis of whether or not their obsessions were triggered by
external cues and whether or not they were associated with catastrophic fears. A theory
which identifies the importance of distinctions like these serves as a guide to the researcher
with regard to the kind of descriptive data to investigate in a focused manner. The fact
that descriptive work is inevitably embedded in theory is shown by considering Giorgis
(1986b, p. 165) examples of questions that might be examined by phenomenological
descriptive research:
How many types of perception are there? What styles of
memorising exist? How varied are neurotic patterns? How are
the relations of love lived out in society?
Concepts such as perception, memory, neurotic patterns and love already carry strong
connotations which are theoretically laden. Ten years later, Wiseman (1995, p.117) asked a
similar question: How many types of loneliness are there?. Her study involved a series of
descriptive phenomenological case studies which enabled her to characterize the subtleties

8
of different experiences of loneliness and to engage creatively with existing
conceptualisations of what it is to be lonely.

PHASE TWO: THEORETICAL-HEURISTIC WORK


In the next phase, descriptions are used as the basis for the generation of or
experimentation with theory (Bromley, 1986; Henwood & Pidgeon, 1993; Strauss & Corbin,
1994). This is the stage of theoretical-heuristic work. This kind of work depends for its
validity on there being a foundation of sound descriptive work, although, as we have
already seen, description and theory building cannot be completely separated in practice.

Grounded theory building


In grounded theory building, theory is built up step by step based on well substantiated
observations. The aim is to develop a conceptualisation, or a framework of distinctions and
relationships, which open up the essential qualities of the type of case being investigated,
and whose value can be re-examined against further cases. Bromley (1986) describes a
similar process which he likens to the building up of case-law in jurisprudence. From the
examination of a few cases of a particular type a basic theory is established which contains
essential concepts, distinctions and principles, linked in a logical manner, which are of
practical value when dealing with such cases. As new cases are brought for examination
some may be dealt with without altering the established case law. Others, however,
require that it be refined, modified or extended. The case-law represents the product of a
progressive conceptual refinement, as knowledge from diverse cases is systematised. At
this stage, care is taken to follow the advice often given to psychotherapists to stay with
the material. This means that the theory remains close to the data and is aptly denoted by
Glaser and Strauss (1967) term "grounded theory" (Henwood & Pidgeon, 1993; Strauss &
Corbin, 1994). What does theory consist of? ask Strauss and Corbin (1994, p 278). It is
useful to think of this kind of theory as having the five components summarised in Table 1.
Each of these will be defined and discussed in turn.

Categorization: The first component is the refining of categories for classification. As we


have seen, this is the interface between focused-descriptive work and theoretical-heuristic
work. The theoretical-heuristic aspect of this is illustrated by Maslows research on
creativity. Long before qualitative researchers were describing the process
metatheoretically, Maslow's work on self-actualization provided clear examples of this kind
of grounded theorising or case law development. Consider this case used to study the
phenomenon of creativity:
One woman, uneducated, poor, a full-time housewife and
mother, did none of these conventionally creative things and
yet was a marvellous cook, mother, wife and homemaker.
With little money, her home was somehow always beautiful.
She was the perfect hostess. Her meals were banquets. Her
taste in linens, silver, glass, crockery and furniture was
impeccable. She was in all these areas original, novel,
ingenious, unexpected, inventive. (Maslow, 1962, p. 136)
This is a case synopsis in which a great deal of behavioural data are summarised and

9
evaluated. Although Maslow's method is implicit (he does not formally present the raw
data or describe the criteria for achieving the data reduction), he could have established
operational criteria for each of these conclusions, had independent judges rate the relevant
behaviours and validate his conclusions, and presented the synopsis more formally. He
uses the case material to identify the need for a conceptual distinction between "special
talent creativeness" and
"self-actualizing (SA) creativeness" which sprang much more
directly from the personality and which showed itself widely in
the ordinary affairs of life. (Maslow, 1962, p 137)
This arose from reflecting on this particular case:
I learned from her and from others that a first rate-soup is
more creative than a second-rate painting (p. 136)
and on others:
It dawned on me once that a competent cellist I had reflexly
thought of as "creative" ... was actually playing well what
someone else had written. (p. 137)

Statements of correlational relationship: A second component of grounded theory is the


identification of correlational relationships (what is associated with what?). Maslows
theorising proceeds to this stage. Having established the criteria for self-actualizing
creativeness he goes on to identify other qualities shown by such people: spontaneity, not
being afraid of ridicule, innocent freedom when creating, and so on.

Definitions of psychological structure: A third kind of theoretical proposition involves


definition of psychological structure. Phenomenological psychologists regard this kind of
work as one of the major tasks of psychology (Giorgi, 1985; Kruger, 1988). This is a distinct
aspect of theory development in which the underlying nature of an experience is
explicated. Brooke's (1985a) study of guilt, using the Duquesne phenomenological method,
illustrates this kind of work. In the first phase of his research, he obtained descriptions
from ten respondents of situations in which they had felt guilty. Synopses of these
(situated structures) were used as the basis for determining a general structure. This is a
phenomenological analysis of the nature of guilt in the form of a set of eleven propositions
about the experience of guilt such that "if any of these constituents are missing, then that
experience is not understood to be one of guilt" (Brooke, 1985a, p. 37). This is a case-based
strategy since the propositions needed to be compatible with the data of every case.
Brookes central proposition about guilt is that it is both ontologically and interpersonally a
rupture between the subjects revealed and private modes of existence, in which a fragile,
apparent harmony and integrity is maintained only by his concealing the fact of his
ruptured existence and the damaged world to which he is indebted. Psychological
structure in this sense is a specialised kind of theorising, informed by a phenomenological
perspective, which goes well beyond mere description. The task of identifying a general
structure does not depend on mechanical steps. Rather it calls for psychological
mindedness, a capacity for insight into the psychological dimensions of the data on the
part of the researcher (Giorgi, 1985).

Similarly, in discussing the task of finding an underlying psychological structure to the event

10
of a person breaking out into laughter, Mahrer (1988, p. 698) recognises that what is called
for is "a particular attitude or perspective which enables what is there is to be discovered
to be seen, and a capacity for creativity and insight that goes beyond the mere
development of conceptual structures for classification. The prerequisites include a good
grasp of the existing knowledge base from the literature, a passive, naive, and patient
stance in which the data are allowed to speak for themselves, an openness to the new, an
eye for the unusual or what does not fit, a willingness to experiment with different ways of
seeing pattern in the data, and an avoidance of the temptation to rush into easy
explanations and overgeneralisations. The researcher is simply advised to "go back to the
data again and again until you receive what is discoverable".
Propositions about process: A fourth component of grounded theory is the identification of
process (what follows from what, and under what conditions?). This is illustrated in the
development of the assimilation model in the Sheffield Psychotherapy Research Project.
Teams of judges listened to the audiotapes of a series of complete, short term
psychotherapies. They progressively refined criteria for classifying eight stages along a
continuum. Three of these stages are: 0) material warded off (but producing symptoms);
3) acknowledges the material and clarifies its existence as a problem that needs to be
tackled; 6) sees how to integrate the material and makes appropriate changes in everyday
life. The categories map a coherent temporal process which forms the basis of the
assimilation model.

Mahrer is describing this kind of theory building when he offers detailed guidelines for
examining the relationship between events within a psychotherapy session. He identifies
three types of event: first, conditions are what is happening with the client at a particular
point in the session (such as becoming tearful, giving a deep sigh, disclosing some intimate
information); second, operations are what the therapist does in response to a condition
(such as making an interpretation or making an empathic remark); third, consequences are
clients' behaviour and experience following the therapist's operations (such as bursting into
hearty laughter or disclosing further information). Research here is guided by a question
such as "If the client self-discloses (condition) and the therapist responds by self-disclosure
(operation) what will the client do (consequence)?" The question is investigated by
obtaining good quality instances of the phenomenon in question, examining them,
preferably with a team of judges, and using an appropriate system of classification.

Propositions about causal relationships: Finally, theory includes propositions about causal
relationships. This means that theory allows for predictability, in the limited sense that if
elsewhere approximately similar conditions obtain, then approximately similar
consequences should occur (Strauss & Corbin, 1994, p 278). Mahrers theorising, just
described, would give rise to this kind of proposition. If the research is successful it would
be possible to predict under what conditions a client would burst into spontaneous
laughter. Many psychologists are responsible for designing interventions of various sorts
that may include individual or group psychotherapeutic procedures, psychoeducational or
training programmes in community development or restructuring or training interventions
within organizations. The evaluation of the effects of such interventions is a central feature
of professional practice, and is often undertaken by means of case studies (Lukoff, Edwards,
& Miller, in press). In such a situation it is important to set up the study to provide

11
conditions under which a causal relationship between the intervention and its effects can
be established. Several points relevant to this are discussed later.

Hermeneutic work
The process just described allows for the development of a coherent discourse within which
refinement of theory can proceed steadily through disciplined investigation and critical
debate. However, a body of theory is negotiated by those who participate in its production.
It is embedded in history (Strauss & Corbin, 1994, p. 280), and inevitably embodies
ontological, epistemological and methodological assumptions which are not tested against
case material (Guba & Lincoln, 1994). This is not problematic, so long as researchers
participating in the process share the same key assumptions, and are examining a well-
defined area of experience. These conditions are largely met within many research
discourses related to practical problems within everyday contexts. This applies to much of
the literature in my own special interest area of cognitive-behaviour therapy. Here, there
are many presentations of rich stores of practical clinical knowledge which are well
grounded in the data of multiple cases. Chambless (1985) discussion of agoraphobia, for
example, or Grayson et als (1985) discussion of obsessive compulsive disorder constitute
presentations of this kind of basic case law (Edwards, 1990; 1996a).

From a post-modern perspective, no discourses or research activities can be so


unproblematic. This perspective emphasises the constructed nature of theoretical
discourses, and the need for an appreciation of the relativity of knowledge structures
(Widdershoven, 1993). This means that any body of grounded theory can be deconstructed
by identifying and critically evaluating the untested assumptions on which it is based (Guba
& Lincoln, 1994). This involves the process of metatheoretical deconstruction, to be
described later. One widely practised solution to this dilemma is for researchers to
consciously situate themselves within the assumptions of a specific discourse for the
purpose of specific research projects. While this allows for systematic and disciplined
theory development, it will inevitably limit the kinds of questions that can be investigated.

In hermeneutic work, researchers appropriate a body of theory not as if it were absolute


truth, but with the recognition that it is historically and culturally constructed. Theory is
conceived of as a lens through which the material can be viewed, and which can provide
access to the deeper dimensions of the case. This approach is of particular value when
investigating a case for which there is not a readily applicable case law. Several existing
concepts or systems of concepts (forestructures) may be appropriated to form the basis of
a reading guide (Kvale, 1996) which is designed to develop an understanding of the case
material. This strategy raises the problem of the hermeneutic circle. The dilemma is that
the framework is used to make sense of the data while the data are, in turn, incorporated
into the framework. Since well developed frameworks often expand to accommodate any
data that may occur, there are questions about the validity of the interpretative process.
Four steps can contribute to overcoming this problem. First, there needs to be a sound
descriptive data base whose validity is established, according to principles to be discussed
later. Second, the argument through which the material is linked to the existing theoretical
discourse needs to be tight and rigorous. Third, as far as possible, the assumptions implicit
in the forestructures should be openly acknowledged. Finally, the case material should be

12
used to critically assess the validity of the conceptual structure of the hermeneutic lenses
being employed. If this can be achieved convincingly, the new case can serve to validate
the grounded nature of the theoretical frames that have been employed.

A hermeneutic case study of my own (Edwards, 1996b) was designed to make several
aspects of this process explicit. A series of past incarnation scenes from the therapy
sessions of a female client were examined. The scenes included being a healer in an
ancient Asclepian Temple, being burned as a witch in the fourteenth century, holding a
dying soldier on a 17th century battlefield, and ministering to wounded and dying soldiers
at a frontline hospital at the end of the 19th century. The first step was to select sessions
related to a common theme (negotiating the nature of spirituality). The second step was to
describe the content of the sessions in a manner which drew out the central features of the
experience but which made no speculative interpretations. Then, in the hermeneutic
phase, four conceptual frames from the literature were employed as hermeneutic keys.
The first, was the process of spiritual emergence (Grof & Grof, 1990), in which various
experiences in altered states of consciousness put a person in touch with deep spiritual
dilemmas. The second was Jungs concept of the collective unconscious, since the
dilemmas portrayed were not purely personal, but collective dilemmas experienced by
many people at many periods of human history. The third was derived from the literature
on regression therapy. This describes past incarnation scenes as one of several experiences
which can occur in altered states of consciousness, and shows how these dramatise
unresolved issues in the personal and collective unconscious. Finally, the core dilemma of
the client was identified which was related to the historical process of repression of the
intuitive and healing qualities of the feminine psyche, as discussed by Metzner (1993).

Each of four concepts, spiritual emergence, collective unconscious, past life regression, and
the repression of the feminine, is situated within a fairly large existing literature. The main
aim of the study was not to build case law or grounded theory in a rigorous manner, but to
use these existing theoretical frames as lenses through which the rather unusual
phenomena of the therapy sessions could be viewed coherently. The success of a
hermeneutic case study depends upon whether the writer can successfully make a case for
the relevance of the hermeneutic frames that have been appropriated. Interested readers
must judge for themselves the extent to which this was achieved. In a further development
of this research, a PhD thesis by Knight (1997) took the form of an extended single case
study, using the same methodology, based on material from later in the therapy of the
same client.

PHASE THREE: THEORY TESTING WORK


The role of the case study in the early phases of the research process has been generally
recognised. Over thirty years ago, Bolgar (1965, p. 31) pointed out:
experimentation is mainly concerned with proof and rarely
leads to discovery ... On the other hand... the case study
method is the ideal way to generate hunches, hypotheses and
important discoveries.
However, in the process of rigorous theory development and the testing of specific
hypotheses case studies can be as important as experimental research, and in some cases

13
more useful (Bromley, 1986; Edwards, 1990: Mitchell 1983; Shapiro, 1970; Stake, 1994,
1995). A well developed case law embodies distinctions and generalisations which can be
empirically tested against new cases. However, in order to test chosen propositions
thoroughly, it will usually be necessary to select a suitable case that provides the conditions
under which crucial tests of theory can be made. In the context of psychotherapy,
discerning clinicians with a well grounded knowledge of existing case law are in a
particularly good position to do original case study work, since they are likely to encounter
such crucial cases from time to time (Mahrer, 1988). The careful management of
conditions, as occurs in single case experimental designs, can also help to create the
conditions required for rigorous theory testing. The fact that, on occasion, a single case
can settle a theoretical point is illustrated in a famous example from mechanics referred to
by Eckstein (1975). Aristotle proposed that the heavier an object the faster it falls. Galileo
argued that weight was irrelevant to rate of fall. In 1650, when it was discovered how to
create a vacuum with an air pump, the single dropping of a coin and a feather was sufficient
to establish that Galileo was right.

Testing propositions within grounded theory


The careful examination of individual cases can be used to critically evaluate and strengthen
or weaken the evidence for propositions relating to any of the five components of
grounded theory discussed above (see Table 1) and to extend existing theory by achieving a
more differentiated body of principles. Some examples will be given by way of illustration,
although no attempt will be made to systematically illustrate the testing of all five
components.

The process of case law development is well illustrated by Rogers, Brown and Tappan
(1994). They conducted a five year longitudinal study of cohorts of girls who, each year,
were interviewed in depth and did a sentence completion test (SCT). They used case
studies to make a detailed examination of the ego development of individual girls, and
made group comparisons of SCT scores obtained in the different years of the study to
examine developmental trends. This approach enabled them to critically examine
propositions within Loevingers theory of ego development. One finding was that
according to the SCT, over a period of three years, very few girls progressed from
conformity to autonomy as predicted by Loevinger, and many actually regressed. The case
studies enabled the researchers to investigate and critically evaluate competing
interpretations of this finding. They noted a paradox: the girls who regressed on the SCT
measure
... struck us as more interesting, intelligent, vital and alive,
more intensely engaged in a struggle for voice and knowledge,
and yet also more likely to reveal signs of psychological
distress (p. 29).
They concluded that these girls had entered a feminine underground in which, on the
surface, they exhibited more conventional and familiar female adolescent roles, while, in
their inner life they were struggling to resist having to live out the gender stereotypes of
passive, silent women (p. 30).

Psychotherapy outcome research is an important example of a process in which hypotheses

14
are examined about correlational or causal relationships between types of event. The
problems with drawing conclusions from single cases about the effectiveness of therapeutic
interventions are well known. Many uncontrolled variables, including extra-therapy events
and non-specific factors in the therapy process itself (client expectations, therapist's
concerned interest) could account for a favourable outcome. It is therefore often argued
that valid conclusions about the effectiveness of psychotherapy interventions can only be
reached by means of a group comparison design in which the critical variables are
measured and controlled. However, the literature on single case experimental design shows
how arguments can be built from data within single cases (Aldridge, 1994, Barker et al,
1994). Kratochwill et al. (1984), building on Kazdin (1981), identified twelve factors which
can strengthen the case for attributing a positive outcome to a particular intervention.
These included failure to respond to previous interventions, the problem being chronic
rather than acute, the therapeutic effect being immediate and large, the use of repeated
objective measurements to assess outcome and the demonstration of the same effect in
new cases, especially if the new cases are unlike the original in some respects. Arguments
can be strengthened when there are data from a series of cases.

Qualitative information about the client's psychological processes can also be evidential,
especially if there is a soundly based case law within which to interpret it. This can be
illustrated from research on the assimilation model in the Sheffield Psychotherapy Project.
Stiles et al. (1991) describe two psychotherapy case studies which illustrate the application
of the model and test some of its principles. The procedure was to identify material from
the therapy dialogue on specific themes and to examine and classify it according to which
stage of the assimilation process was evidenced in each session. In accordance with the
theory, material was found to move along the continuum from being less assimilated to
being more assimilated as the therapy progressed. The data therefore support the main
tenet of the assimilation model. However, Stiles et al. (1991, p. 203) comment that
selected case studies do not test the model but serve only to
illustrate this qualitative approach... Formal testing requires
derivation of hypotheses that can be accepted or rejected.
This is only partly true. If a model if derived from one or two cases, then those original
cases provide an empirical basis for the statement of the model but new cases are needed
to test it. Since the cases presented by Stiles et al are not the original cases used to derive
the model, they do test the model and do not merely illustrate it. Stiles et al. may mean
that the cases have been specially selected for presentation because they fit the model,
whereas some other cases which fail to fit have not been presented. Even if this is the case
the selected cases still support the theory. Other cases which fail to fit the theory would
not necessarily result in the rejection of a conclusion which has been established on the
basis of many cases (provided that the conclusion was well grounded in the data in the first
place). Rather they might be expected to result in extension or refinement by raising such
questions as "Under what conditions does material on a specific theme fail to become
assimilated over a series of therapy sessions", or "Under what conditions does
unassimilated material on a theme appear when there is other evidence that the theme is
already well assimilated?"

Within the field of psychotherapy, this understanding of the nature of case law provides a

15
means for therapists to contribute to the research process through work with routine cases
(Barker et al, 1994; Edwards, 1990, 1996a). In working, for example, with cases of obsessive
compulsive disorder, therapists need a thorough knowledge of the existing literature and
need to situate and conduct the therapy with reference to that literature. If the case
conceptualisation, therapeutic interventions and outcome can be understood entirely in
terms of the existing case law, it will be a working case study. In providing an example of
the successful application of existing theory, the case serves to validate that theory.
However, if the material of the case requires that a new phenomenon be recognised, new
distinctions made, or a different relationship between phenomena be recognised, it can
provide the basis for a formal exercise in theory development. The interested reader is
referred to Salkovskis (1985; summarized in Edwards, 1990 & 1996a) for an example of this
approach in practice.

Mahrer also discusses how theoretical propositions and hypotheses are first formulated
from a set of relevant cases and then tested them against new ones. This process is the
basis for "rigorous scientific theory building" (Mahrer, 1988, p 701). He points out that
although such hypotheses could be examined in quantitative group comparison designs, the
intensive analysis of series of cases can be a more efficient and informative means of
refining, and testing conceptual frameworks and theoretical propositions. As such it is not
simply "a somewhat useful secondary tool for the serious work of scientific hypothesis
testing" (Mahrer, 1988, p. 697) but a substantial method for advancing knowledge about
psychotherapy, that will, he believes, eventually "unseat hypothesis testing as the scientific
means of enquiry in the field of psychotherapy research" (p. 701).

Step four of Mahrers discovery oriented research model provides an example of the testing
of hypotheses about psychological structure. Here, new instances are gathered of the
phenomenon being investigated, and they are used to test the generality of the conceptual
framework already derived. Brooke employed the same step in his study of guilt. Having
developed his account of the structure of guilt on the basis of ten cases, he proceeded to
test the validity of each proposition against fifty-two other cases. In the third phase, this
analysis formed the basis for a critical examination of theoretical statements about the
nature of guilt derived from Freud (Brooke, 1985a) and Jung (Brooke, 1985b). He observed,
for example, that his material validated Freud's view that guilt can be "unconscious",
although he reframed this as "an ambiguous consciousness of guilt" whereby "a person can
live and experience guilt, even painfully, without knowing that it is guilt that he feels, or
why" (Brooke, 1985a, p. 43).

Metatheoretical deconstruction
Metatheoretical deconstruction involves using material from cases to expose the hidden
assumptions on which a body of psychological theory is based. The consequence of this is
not the refining of theory, but a more radical reformulation. Not all deconstruction is
metatheoretical in this sense. For example, many investigations using discourse analysis
deconstruct the assumptions implicit in the language of everyday conversations or
newspaper articles (Levett, Kottler, Burman & Parker, 1997). This is effectively a form of
hermeneutic work (as discussed above) in which a body of psychological theory is used to
deepen understanding of a text.

16
Metatheoretical deconstruction of positivist or dualist conceptions of human nature is
frequently undertaken on the basis of the findings of phenomenological research studies.
This is illustrated in Brookes study, where he used his data on guilt to criticize Freuds
dualistic separation of individual and world and his quasi-spatial topographical analysis of
the psyche. He argued that None of this... is compatible with an articulation of man
understood existentially, as being-in-the-world (p. 42). Only by deconstructing Freuds
metapsychology can he initiate a dialogue between Freuds understanding of guilt and his
research findings. Similarly, Stevick (1971), used her findings on the phenomenology of
anger to metatheoretically deconstruct the assumptions implicit in Schachters classic
studies. She concluded that emotion is a response of the whole person to a situation,
which is lived before it is reflected on ... arousal is the bodys affective mode and argued
that this understanding is incompatible with Schachters assumption that emotion can be
reduced to cognition and responses of the body as merely a physiological organism (p
145). In concluding her study of the immigrant experience, Polyzoi (1985) engaged in a
similar kind of deconstructive work. She criticized as too limiting, the traditional view of the
immigrants assimilation into a new society which focuses on the how the immigrant
successively approximates the appearance, behaviour, language and cultural values held by
members of the receiving society (p 66). She suggested that in order to adequately
accommodate the complexities of the experience found in her study there is a need to
create a more adequate language to address this complex experience (p. 68).

There is inevitably a continuum between what we have already described as the process of
testing hypotheses about psychological structure and the process of metatheoretical
deconstruction. However, the basis of metatheoretical deconstruction is that the
researcher adopts a perspective which is radically different in terms of assumptions from
those of the theory being deconstructed. Stevicks conclusion could be seen as leading to
an enrichment of existing case law by adopting a phenomenological stance. In this case the
work would be an examination and extension of theories of psychological structure.
However, as a metatheoretical deconstruction, it could lead to a more radical conclusion: a
reformulation which effectively replaced existing constructs.

VALIDITY IN CASE STUDY RESEARCH


To be rigorous, case-based researchers need to attend to issues of validity at every step. An
account of the principles for establishing validity and reliability in quantitative research has
become an established part of research methodology teaching. There has been a similar
longstanding appreciation of issues of reliability and validity in the literature on case-based
research (Bromley, 1986; Polkinghorne, 1986; M. B. Shapiro, 1966, 1970, 1979; Taylor &
Bogdan, 1984; Wertz, 1986; Yin, 1984). Several recent discussions are comprehensive and
conceptually well differentiated (Kvale, 1996; Miles & Huberman, 1994).

Much case study research has been rightly criticized because of undisciplined writing in
which there was inadequate concern for validity (Rosal, 1989). Case studies lack rigour and
are methodologically flawed if they use unreliable self-report data, unsubstantiated
observations, post-hoc and unsystematic therapist summaries of the content of therapy
sessions, or when they lapse into speculation and overgeneralisation in drawing

17
conclusions. These points are contained in Strupps (1981, pp. 217-8) damning critique of
the application of case study method in the psychoanalytic movement:
total commitment to a single theoretical and technical model;
enshrinement of theoretical formulations as dogma; imper-
viousness to research progress in neighbouring fields; rigid
indoctrination of trainees; failure to inspire students to
become critical and independent thinkers; failure to teach
students abiding respect for empirical data and rival
hypotheses; insufficient attention to clinical realities.

Like all research, a case-based research project proceeds in a step by step manner and the
value of each step depends on the validity of the preceding steps. At each step, there are
threats to validity, which need to be anticipated and taken care of methodologically.
Whether the aim of a project is descriptive, theoretical-heuristic or theory-testing, case
studies will involve technical procedures for data collection, data reduction and data
interpretation (Miles & Huberman, 1994). During data collection, steps must be taken to
ensure that the raw data are genuine and free from bias and selection effects. Awareness
of the threats to validity of data gathered by means of interviews is not new (Hyman, 1954;
M. B. Shapiro, 1970, 1979). It is well known that interviewers need to know how to avoid
asking leading questions, activating response sets, setting up situations with strong
demand characteristics, or selectively recording responses. Similarly, where data include
other forms of self-report (or collateral reports from others) steps must be taken to assess
the extent to which information can be relied on.

In the Duquesne tradition, as in other phenomenological approaches, there is concern for


the validity of the data as an accurate representation of the interviewee's experience. Data
collection is to some extent modelled on the manner in which psychotherapists develop an
understanding of clients' unfolding experience in psychotherapy. Like clinicians,
interviewers work to create a context in which a relationship of trust can develop. They
strive to attune to the interviewee's world by empathic listening and skillful questioning
and to suspend personal judgements and preconceptions regarding what is described, so
that interviewees will disclose psychologically meaningful information.

Data reduction is the process whereby a large and cumbersome body of data (often in the
form of long transcripts of interviews, sometimes from several different people), is
organised into a manageable form both for the researcher to work with and for
presentation. The concern here is that case histories, case synopses or interview
summaries faithfully portray the experience and behaviour of the participant and are free
from selection bias (omission of aspects of the material which might be problematic for the
researchers favoured assumptions). The brilliant and pioneering psychotherapy case
studies of Freud served as the foundation for the development of psychoanalytic theory
(Bolgar, 1965). But studies like these have been criticised for their failure to secure validity
at this stage. Enthusiasm for a particular interpretation, or desire to make a particular
point can lead to writing case synopses which are distorted by the aims and assumptions of
the writer. Spence (1986) refers to this as narrative smoothing. Barker et al (1994, p.
166) suggest this experiment: tape-record a therapy session, but do not take notes; two

18
days later, write a synopsis of what happened in the session; listen to the tape-recording.
They claim that,
In addition to large amounts of missing material, you will also
find that you have collapsed things that happened at different
times, got some things out of order, may have attributed
statements to the wrong speaker and even completely
fabricated things.
These problems can be overcome by using tape-recordings, having independent judges
listen to or read the transcript and evaluate the synopses made from them. Another check
is to have participants themselves read the case synopses and indicate whether their
experience has been faithfully portrayed.

During data interpretation, in which new theory is developed or aspects of existing case law
tested, validity depends on the quality of the arguments that link data to theory. The
construction of a rational and persuasive chain of reasoning building from data to
conclusions is called "analytic generalisation" by Yin (1984 p. 39) and "analytical induction"
by Mitchell (1983, p. 200). M. B. Shapiro (1970, p. 664), who recognized that both
induction and deduction are involved, refers to an inductive-deductive approach and
emphasises the method of the systematic investigation as a means of improving the
validity of our conclusions (p. 651). Once hypotheses have been proposed, alternative
explanations and theories need to be canvassed and tested. At this stage, we see the
strength of in-depth case studies in which there is a wealth of data systematically collected
from multiple sources, since detailed examination of the data can provide a basis for
critically evaluating different interpretations and conclusions within a single case (Barker et
al., 1994). Where there is insufficient information to establish a particular conclusion, it
may be possible to seek further data from the same case that can cast light on the
researchers question (Shapiro, 1970).

Another important strategy here is to examine explanations and interpretations across


cases. Hypotheses are best secured if they are tested against a broad range of cases.
Researchers should actively search for "test cases" (Bromley, 1986, p. 26) or negative
cases (Henwood & Pidgeon, 1993) which enable competing theories to be evaluated
against each other or that might clearly disprove a particular hypothesis (Miles &
Huberman, 1994; Taylor & Bogdan, 1984).

These strategies can be applied to the establishing of internal validity. This refers to any
claim that a causal relationship has been demonstrated. Some more specific strategies and
arguments have been discussed above with specific reference to the internal validity of
claims about the causes of outcome in psychotherapy. External validity refers to the
generalisability of conclusions from a research setting to real situations. Because case
studies examine people in or close to real situations and do not distort naturally occurring
behaviour through experimental manipulation and the setting up of artificial conditions,
they usually have better external validity than controlled experiments (Kiesler, 1981; Taylor
& Bogdan, 1984; Yin, 1984). Indeed this is one the strengths of the well conducted case
study.

19
CASE-BASED RESEARCH: SOME COMMON PITFALLS
The framework described here has been developed out of the experience of the difficulties
students have in implementing the case-based research process. Supervisors and students
may find it valuable to ask the question: What kind of work am I engaged in at this stage of
the research process, and (for the later phases) can this kind of work be adequately
sustained by the technical approach I have chosen and by the results of the kinds of work I
have undertaken in previous phases of the study. In addition to helping researchers to
focus on the issues of validity discussed above, these questions can highlight the nature of a
number of serious errors students often make.

The first is the token literature review. Here, the student reviews the literature but fails to
conceptualise or evaluate the state of existing case law, or ignores existing case law in
formulating the research question, designing the method or discussing the results. These
failures completely flaw case studies methodologically since it is the essence of case-based
research that new cases contribute to knowledge only in the context of the case law
developed from previous cases.

Another frequent error is premature theorising: the researcher embarks on theory


development without establishing a valid descriptive base to support it. Related to this is a
set of errors due to the pseudo qualitative fallacy. Here, the students research proposal
provides the basis for a case-based investigation of respondents personal meanings, but
the students hidden agenda is to draw the kinds of conclusions which can only be drawn
using a quantitative approach or a group comparison design. One form of this is to draw
general conclusions on the basis of a comparison between only two cases. In its grossest
form, a male and female are studied and conclusions are drawn about men and
women. Another form is to employ a kind of informal quantification in summarizing the
material from several cases. The researcher writes, for example a few individuals
experienced ..., while others said ..., however most .... This could be of value as a
preliminary step in surveying the kinds of experiences which respondents report. It could
be a preamble to a systematic classification of kinds of experience, which could, in turn,
give rise to a formal quantitative investigation of the frequency of each kind of experience
in a population. Unfortunately, students often try to draw conclusions about population
frequencies from this kind of informal preliminary data.

Phase slippage is a methodological problem that occurs when researchers fail to separate
out the different stages of the research process. Here, the phases of the research process
are not presented in a systematic order, or, even if they are, material which belongs in one
phase is introduced during another phase. An anthropological paper which I came across
recently exemplifies this problem. In this fascinating account of Tamazai, a culture-bound
disorder among the Tuareg (Rasmussen, 1992), the author, moves freely between
descriptions of the disorder, informants theories about the disorder, her own hermeneutic
reflections and theoretical speculation. As a result it is difficult for readers not already
deeply familiar with the field to gain a sense of coherent argument. The interested reader
is invited to read Rasmussens paper and reflect on it from the perspective of the principles
set out here. Where this occurs in students work, this error is likely to be disastrous since
they seldom have an underlying thorough grasp of the material in the way that an

20
experienced anthropologist like Rasmussen has. In supervising students, therefore, it is
particularly important to help them to make explicit the different steps of the research
process and to be mindful of the kind of case study work that is being done at each step.

CONCLUSIONS
One aim of this paper has been to correct commonly expressed misconceptions about case-
based research. In a textbook on child psychopathology, for example, Gelfand, Jenson and
Drew (1988, p. 323), state that case studies are of no value because they cannot be used to
derive general principles. This is a serious misrepresentation. It is the examination of a
series of cases that forms the basis of the systematic construction of a body of knowledge,
and as the literature in a specific area develops, the researcher who studies a single case in
depth can draw on a corpus of similar previously reported cases as the basis for the
examination and development of theory.

Another aim is to encourage intensive research with single cases. Except perhaps in the
area of psychobiography (Denzin, 1989; Smith, 1994), present models, even in qualitative
research, tend to discourage students from projects involving depth analysis of single cases.
The framework presented here has served to inform the supervision of two recent PhD
theses each of which examined the material from a single psychotherapy case (Knight,
1997; Oberholzer, 1996).

It is hoped that this paper will encourage researchers to contribute to the process of
ongoing theory development through the investigation of individual cases. This has always
been, explicitly or implicitly, at the core of most theory development in clinical psychology,
and in many areas of personality theory and developmental psychology. The principles set
out here relieve case study research of its status as a second class citizen and allow it to be
appreciated as a fundamental tool for the development of valid knowledge in the human
sciences, or, to use Bromley's (1986, p. ix) phrase, "the bedrock of scientific investigation".

Footnotes

1. This research has been supported by grants from: the Centre for Science
Development, the Centre for Research Methodology, and the Rhodes University
Joint Research Committee.

2. I am grateful to Ann Shuttlewoth-Jordan for her valuable comments on an earlier


draft of this paper.

3. I am grateful to David Lukoff for providing the material of this paragraph as well as
for his encouragement and support for the writing of the paper as a whole.

21
References
Aldridge, D. (1994). Single case research designs for the creative art therapist. The Arts in
Psychotherapy, 21, 333-342.
Barker, C., Pistrang, N., & Elliot, R. (1994). Research methods in clinical and counselling psychology.
Chichester: John Wiley.
Beck, A.T., Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cognitive therapy of depression. New York:
Wiley.
Beck, A.T. & Emery, G. (1985). Anxiety disorders and phobias: A cognitive perspective. New York:
Basic Books.
Barlow, D. H. (1981). On the relation of clinical research to clinical practice: Current issues, new
directions. Journal of Consulting and Clinical Psychology, 49, 147-155.
Bolgar, H. (1965). The case study method. In B. B. Wolman (Ed.) Handbook of clinical psychology
(pp.28-39). New York: McGraw Hill.
Bromley, D. B. (1986). The case-study method in psychology and related disciplines. Chichester:
Wiley.
Brooke, R. (1985a). What is guilt? Journal of Phenomenological Psychology, 16, 31-46.
Brooke, R. (1985b). Jung and the phenomenology of guilt. Journal of Analytical Psychology, 30,
165-184.
Chambless, D.L. (1985). Agoraphobia. In, M. Hersen & A. S. Bellack (Eds.). Handbook of clinical
behavior therapy with adults (pp. 49-88). New York: Plenum.
Cleaver, G. (1988). A phenomenological analysis of victimization. The experience of having one's
house attacked and damaged. South African Journal of Psychology, 18, 76-83.
de Koning, A. J. J. (1979). The qualitative method of research in the phenomenology of suspicion.
In, A. Giorgi, R. Knowles & D. L. Smith (Eds.), Duquesne studies in phenomenological psychology
(Volume 3) (pp.122-134). Pittsburgh: Duquesne University Press.
Denzin, N. K. (1989). Interpretive biography. Newbury Park: Sage.
Eckstein, H. (1975). Case study and theory in political science. In, F. I. Greenstein & N. W. Polsby
(Eds.), Handbook of political science: Volume 7 - Strategies of enquiry (pp.79-137). Reading,
Mass.: Addison-Wesley.
Edwards, D. J. A. (1990) Research and reality: How clinical theory and practice are actually
developed - Case study method in cognitive-behaviour therapy. In, J. Mouton & D. Joubert,
(Eds.), Knowledge and method in the human sciences (pp.359-374). Pretoria: Human Sciences
Research Council.
Edwards, D. J. A . (1991) Duquesne phenomenological research method as a special class of case
study research method. In R. van Vuuren, (Ed.), Dialogue beyond polemics (pp.53-70). Pretoria:
Human Sciences Research Council.
Edwards, D. J. A. (1996a). Case study research method: The cornerstone of theory and practice. In,
M. Reinecke, F. Dattilio & A. Freeman (Eds.). Cognitive therapy with children and adolescents: A
casebook for clinical practice (pp.10-37). New York: Guilford.
Edwards, D. J. A. (1996b). The old gods are awakening: A hermeneutic case study of spiritual
emergence. In, C. R. Stones (Ed.), Qualitative research approaches to psychotherapy process
(pp 101-122). Pretoria: Human Sciences Research Council.
Erikson, E. H. (1969). Ghandis truth: On the origins of militant non-violence. New York: W. W.
Norton.
Erikson, E. H. (1993). Young man Luther: A study in psychoanalysis and history. New York: W. W.
Norton.
Fischer, W. F. (1982). An empirical-phenomenological approach to the study of anxiety. In, A. J. J. de
Koning & F. A. Jenner (Eds.), Phenomenology and psychiatry (pp.63-84). London: Academic.
Gelfand, D. M., Jenson, W. R. & Drew, C. J. (1988). Understanding child behavior disorders: An

22
introduction to child psychopathology. Fort Worth: Holt, Rinehart & Winston.
Giorgi, A. (1985). Sketch of a psychological phenomenological method. In A. Giorgi (Ed.),
Phenomenology and psychological research (pp. 82-103). Pittsburgh: Duquesne University
Press.
Giorgi, A. (1986a). Theoretical justification for the use of descriptions in psychological research. In
P. D. Ashworth, A. Giorgi & A. J. J. de Koning (Eds.), Qualitative research in psychology:
Proceedings of the International Association for Qualitative Research in Social Science (pp.
3-22). Pittsburgh: Duquesne University Press.
Giorgi, A. (1986b). The "Context of discovery/context of verification" distinction and descriptive
human science. Journal of Phenomenological Psychology, 17, 151-166.
Glaser, B. G. & Strauss, A. L. (1967). The discovery of grounded theory. London: Weidenfeld &
Nicolson.
Grayson, J. B., Foa, E., & Steketee, G. (1985). Obsessive-compulsive disorder. In M. Hersen, & A. S.
Bellack (Eds.), Handbook of clinical behavior therapy with adults (pp. 133-165). New York:
Plenum.
Greenberg, L. S., & Pinshof, W. M. (1986). The psychotherapeutic process: A research handbook.
New York: Guilford.
Grof, C., & Grof, S. (1990). The stormy search for the self. Los Angeles: Tarcher.
Guba, E. G., & Lincoln, Y. S. (1994). Competing paradigms in qualitative research. In N. K. Denzin &
Y. S. Lincoln (Eds.), Handbook of qualitative research (pp. 105-117). Thousand Oaks, CA: Sage.
Hayes, S. C. (1981). Single case experimental design and empirical clinical practice. Journal of
Consulting and Clinical Psychology, 49, 193-211.
Henwood, K. L., & Pidgeon, N. F. (1993). Qualitative research and psychological theorising. In M.
Hammersley (Ed.), Social research: Philosophy, politics and practice (pp.14-32) . London: Sage.
Hyman, H. H. (1954). Interviewing in social research (Reprinted, 1975). London: University of
Chicago Press.
Josselson, R. & Lieblich, A. (Eds.) (1993). The narrative study of lives, Volume 1. Newbury Park: Sage.
Josselson, R. & Lieblich, A. (Eds.) (1995). The narrative study of lives, Volume 3: Interpreting
experience. Thousand Oaks: Sage.
Kazdin, A.E. (1981). Drawing valid inferences from case studies. Journal of Consulting and Clinical
Psychology, 49, 183-192.
Kiesler, D.J. (1981). Empirical clinical psychology: Myth or reality? Journal of Consulting and Clinical
Psychology, 49, 212- 215.
Knight, Z. (1997). Healing stories of the unconscious: Past life imagery in transpersonal
psychotherapy. Unpublished PhD thesis, Rhodes University, Grahamstown, South Africa..
Kratochwill, T. R., Mott, S. E., & Dodson, C. L. (1984). Case study and single-case research in clinical
and applied psychology. In A. S. Bellack & M. Hersen (Eds.), Research methods in clinical
psychology (pp.55-99). New York: Pergamon.
Kruger, D. (1988). An introduction to phenomenological psychology (2nd ed.). Cape Town: Juta.
Kvale, S. (1996). InterViews: An introduction to qualitative research interviewing. Thousand Oaks
CA: Sage.
Levett, A., Kottler, A., Burman, E., & Parker, I. (1997). Culture, power and difference: Discourse
analysis in South Africa. London: Zed Books.
Lieblich, A. (1993). Looking at change: Natasha, 21: New immigrant from Russia to Israel. In R.
Josselson, & A. Lieblich (Eds.), The narrative study of lives, Volume 1 (pp. 92-129). Newbury
Park: Sage.
Lieblich, A., & Josselson, R. (Eds.) (1994). The narrative study of lives, Volume 2: Exploring identity
and gender. Thousand Oaks: Sage.
Lukoff, D. (1985). The diagnosis of mystical experiences with psychotic features. Journal of

23
Transpersonal Psychology, 17, 155-181.
Lukoff, D., Lu, F. & Turner, R. (1998). From Spiritual Emergency to Spiritual Problem. The
transpersonal roots of the new DSM-IV category. Journal of Humanistic Psychology, 38(2), ***-
***.
Lukoff, D. & Everest, H.C. (1985). The myths of mental illness. Journal of Transpersonal
Psychology, 17, 123-153.
Lukoff, D., Edwards, D., & Miller, M. (In press). The case study as a scientific method for studying
alternative therapies. Alternative Therapies in Health and Medicine.
Mahrer, A.R. (1988). Discovery-oriented psychotherapy research: Rationale and aims. American
Psychologist, 43, 694-702.
Maslow, A. (1962). Towards a psychology of being. New York: van Nostrand.
Maslow, A. (1973). The farther reaches of human nature. Harmondsworth: Pelican.
May, R. (1942). Power and innocence. New York: Norton.
Merleau-Ponty, M. (1967). The structure of behavior. Boston: Beacon.
Metzner, R. (1993). The split between spirit and nature in European consciousness. ReVision, 15,
177-185.
Miles, M. B. & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd
ed.). Thousand Oaks CA: Sage.
Mitchell, J. C. (1983). Case and situation analysis. Sociological Review, 31(2), 187-211.
Montgomery, J. D. (1982). A phenomenological investigation of the premenstruum. Journal of
Phenomenological Psychology, 13, 45-72.
Oberholzer, S. A. (1996). Breaking the karmic complex: The role of transpersonal phenomena in
psychotherapy with an adult survivor of child abuse. A clinical case study. Unpublished PhD
thesis, Rhodes University, Grahamstown, South Africa.
Polkinghorne, D. E. (1986). Conceptual validity in a non-theoretical human science. Journal of
Phenomenological Psychology, 17, 129-150.
Polyzoi, E. (1985). Reflective phenomenology: An alternative approach to the study of the
immigrant experience. Journal of Phenomenological Psychology, 16, 47-72.
Rasmussen, S. J. (1992). Reflections on Tamazai: A Tuareg idiom of suffering. Culture, Medicine and
Psychiatry, 16, 337-365.
Reason, P. (1988). Human inquiry in action: Developments in new paradigm research. London:
Sage.
Rice, L. & Greenberg, L. S. (1984). Patterns of change: Intensive analysis of psychotherapy process.
New York: Guilford.
Rogers, A. G., Brown, L. M., & Tappan, M. B. (1994). Interpreting loss in ego development in girls.
In A. Lieblich, & R. Josselson (Eds.), The narrative study of lives, Volume 2: Exploring identity
and gender (pp.1-36). Thousand Oaks: Sage.
Rogers, C. (1942). Counseling and psychotherapy. Boston: Houghton Mifflin.
Rosal, M. L. (1994). Single case research designs for the creative art therapist. The Arts in
Psychotherapy, 16, 71-75.
Rosenthal, G. (1993). Reconstruction of life stories: Principles of selection in generating stories for
narrative biographical interviews. In R. Josselson, & A. Lieblich (Eds.), The narrative study of
lives, Volume 1 (pp. 59-91). Newbury Park: Sage.
Salkovskis, P.M. (1985). Obsessional-compulsive problems: A cognitive-behavioural analysis.
Behaviour Research and Therapy, 23(5), 571-583.
Safran, J., Greenberg, L. S., & Rice, L. (1988). Integrative psychotherapy research and practice:
Modeling the change process. Psychotherapy, 25, 1-17.
Shapiro, M. B. (1966). The single case in clinical psychological research. Journal of General
Psychology, 74, 3-23.

24
Shapiro, M. B. (1970). Intensive assessment of the single case: An inductive-deductive approach. In
P. Mittler (Ed.), The psychological assessment of mental and physical handicaps (pp. 645-666).
London: Methuen.
Shapiro, M. B. (1979). Assessment interviewing in clinical psychology. British Journal of Social and
Clinical Psychology, 18, 211-218.
Smith, L. M. (1994). Biographical method. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of
qualitative research (pp. 286-305). Thousand Oaks, CA: Sage.
Spence, D. P. (1986). Narrative smoothing and clinical wisdom. In T. R. Sarbin (Ed.), Narrative
psychology: The storied nature of human conduct. New York: Praeger.
Stake, R. E. (1994). Case studies. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative
research (pp. 236-247). Thousand Oaks, CA: Sage.
Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.
Stiles, W. B., Elliot, R., Llewelyn, S. P., Firth-Cozens, J. A., Shapiro, D. A. & Hardy, G. (1990).
Assimilation of problematic experiences by clients in psychotherapy. Psychotherapy, 27, 411-
420.
Stiles, W. B., Morrison, L. A., Haw, S. K.. Harper, H., Shapiro, D. A. & Firth-Cozens, J. (1991).
Longitudinal study of assimilation in exploratory psychotherapy. Psychotherapy, 28, 195-206.
Stevick, E. L. (1971). An empirical investigation of the experience of anger. In, A. Giorgi, W. F.
Fischer, & R. von Eckartsberg, R. (Eds.). Duquesne studies in phenomenological psychology (Vol.
1) (pp.132-148). Pittsburgh: Duquesne University Press.
Strauss, A., & Corbin, J. (1994). Grounded theory methodology. In N. K. Denzin & Y. S. Lincoln (Eds.),
Handbook of qualitative research (pp. 273-285). Thousand Oaks, CA: Sage.
Strupp, H. H. (1981). Clinical research, practice, and the crisis of confidence. Journal of Consulting
and Clinical Psychology, 49, 216-219.
Taylor, S. J. & Bogdan, R. (1984). Introduction to qualitative research methods: The search for
meanings. (2nd ed.). New York: Wiley.
Thorpe, M. R. (1989). A phenomenological investigation into the psychotherapist's experience of
identifying, containing and processing the patient's projective identification. Unpublished Ph.D.
thesis. Rhodes University, Grahamstown, South Africa.
Todres, L. A. (1990). An existential-phenomenological study of the kind of self insight that carries a
greater sense of freedom. Unpublished Ph.D. thesis. Rhodes University, Grahamstown, South
Africa..
Weiland, S. (1995). Life history and academic work: The career of Professor G. In R. Josselson, &
A. Lieblich (Eds.), The narrative study of lives, Volume 3: Interpreting experience (pp. 59-99).
Thousand Oaks: Sage.
Wertz, F. J. (1985). Method and findings in a phenomenological psychological study of a complex
life-event: Being criminally victimized. In, A. Giorgi (Ed), Phenomenology and Psychological
Research (pp.155-216). Pittsburgh: Duquesne University Press.
Wertz, F. J. (1986). The question of the reliability of psychological research. Journal of
Phenomenological Psychology, 17, 181-205.
White, R. (1972). Lives in progress: A study of the natural growth of personality. Henry Holt.
Widdershoven, G. A. M. (1993). The story of life: Hermeneutic perspectives on the relationship
between narrative and life history. In R. Josselson, & A. Lieblich (Eds.), The narrative study of
lives, Volume 1 (pp. 1-20). Newbury Park: Sage.
Wiseman, H. (1995). The quest for connectedness: Loneliness as process in the lives of lonely
university students. In R. Josselson, & A. Lieblich (Eds.), The narrative study of lives, Volume 3:
Interpreting experience (pp. 116-152). Thousand Oaks: Sage.
Yin, R.K. (1984). Case study research: Design and methods. Beverly Hills: Sage.

25

Das könnte Ihnen auch gefallen