Sie sind auf Seite 1von 12

5~. .Scl. rfzd. Vol. 30. No. I I. pp. 1161-l 172.

1990 0277-9536190
53.00+ 0.00
Pnnted in Grca! Bntaln. ,411rights resrned Copyright i” 1990Pergamon Press plc

‘DISCOVERING’ CHRONIC ILLNESS: USING GROUNDED


THEORY _
KATHY CHARMAZ
Sonoma State University, Rohnert Park, CA 94928, U.S.A.

Abstract-This paper focuses on using the grounded theory method to study social psychological themes
which cut across diverse chronic illnesses. The grounded theory method is presented as a method having
both phenomenoiogical and positivistic roots, which leads to confusion and misinterpretations of the
method. A social constructionist version and application of grounded theory are introduced after brief
overviews of the method and of the debates it has engendered are provided. Next, phases in developing
concepts and theoretical frameworks through using the grounded theory approach are discussed. These
phases include: (I) developing and refining the research and data collection questions, (2) raising terms
to concepts. (3) asking more conceptual questions on a generic level and (4) making further discoveries
and clarifying concepts through writing and rewriting. Throughout the discussion, examples and
illustrations are derived from two recent papers, ‘Disclosing Illness’ and ‘Struggling for a Self: Identity
Levels of the Chronically III’. Last, the merits of the method for theoretical development are discussed.

ISTRODUCTION the meanings that people ascribe to their situations


[8-l I]. These meanings derive from shared inter-
The grounded theory method provides a set of useful actions, which turn on the pivotal role of language.
research strategies for studying the experience of The symbolic interactionist assumption of the
chronic illness [I-5]. This paper focuses on using the indeterminancy of action rests on the human capacity
grounded theory method for studying people with to objectify self and to ascribe meanings to self like
a wide range of chronic illnesses to address how any other object. Such a perspective can lead to an
chronicity affects ill people’s self-concepts. Through- overly rationalized view of the individual which
out the paper, ways of using grounded theory to a phenomenological perspective helps to correct.
develop social constructionist analyses are explored. Because phenomenology means studying the objects
As used here, the term social constructionist means: of consciousness, it fosters studying emotions [ 12, 131.
(I) III people’s creation of taken-for-granted inter- Both symbolic interactionism and phenomenology
actions, emotions, definitions, ideas, and knowledge lead the researcher to look closely at the research
about illness and about self and (2) Researchers’ participants’ interpretations of their actions and
sociological constructions which they develop, in situations. Marxist theory can then provide tools for
turn, by studying chronically ill people’s construc- linking subjective consciousness and choice to larger
tions. social structures. Moreover, Marxist theory brings a
Chronically ill people, like most everyone, experi- critical posture to examining the data.
ence their constructions as reality; their constructions For the type of social constructionist view taken
are neither convenient fabrications nor idiosyncratic here, the assumptions underlying these theoretical
inventions. Rather, ill people’s constructions reflect perspectives and the questions flowing from them are
their understandings of their experiences as well as perhaps, more significant for the resulting grounded
the diverse situations in which they have them [6]. theory analyses then are specific concepts inherent
Further. their friends and family often support within each theoretical perspective. Symbolic inter-
their constructions even when these constructions actionists assume that as thinking, acting, creative
challenge or contradict those of medical profession- individuals, human beings respond to the actions of
als, and even when ill people cannot make their others after interpreting these others’ intent and
constructions credible or negotiable. Grounded action. A symbolic interactionist perspective leads
theory analyses can then provide physicians with one to look at self and meaning as processes.
alternative understandings of patients’ beliefs and Phenomenologists assume that subjective reality may
actions than those readily available in clinical take varied forms. This perspective fosters the re-
settings. Subsequently, physicians may use these searcher’s study of the multiple dimensions and
understandings to improve communications with realities of a person’s lived experience. Marxists
patients and to act on problems which patients define. assume that individual psychology is both shaped and
Like Foucault’s [7] argument about the conditions constrained by social structure. Thus, a Marxist
giving rise to discursive practices, this constructionist perspective fosters asking critical questions about
view acknowledges that outcomes result from social how society impinges upon the individual and how
interactions, negotiations and power. However, individuals reproduce dominant ideas within society.
my application of grounded theory derives from a In keeping with Berger and Luckman [14], such a
symbolic interactionist perspective tempered by constructionist view assumes an emergent reality
Marxism and phenomenology. Symbolic inter- fundamentally shaped by social interaction. Hence, a
actionism assumes that human action depends upon constructionist approach offers an open-ended and

1161
1162 KATHY CHARMAZ

flexible means of studying both fluid interactive pro- researcher’s approach to data collection and analysis.
cesses and more stable social structures [15]. Rather In comparison, more traditional logical-deductive
than directly applying concepts such as Mead’s “I”, approaches explicitly derive hypotheses from pre-
“me”, and “generalized other”, or Schutz’ “typi- existing theories, which fundamentally structure both
fication” or “multiple realities”, these concepts were the data collection and analysis toward verification of
used to sensitize me to look for themes and issues refutation of these hypotheses (and therefore, the
within the data [16]. theories from which they were derived). Grounded
Themes were addressed which cut across diverse theorists affirm. check. and refine their developing
chronic illnesses such as multiple sclerosis, diabetes, ideas, but they do not limit themselves to pre-
circulatory disease, renal failure, and cancer. These conceived hypotheses nor do they follow the pre-
cross-cutting themes initially included self-esteem, scribed canons of traditional random sampling
continuity and change of self concept, and relation- required for statistical verification.
ships between time and identity. After gathering and Grounded theory also differs from other qualita-
studying more data, my focus also explicitly included tive approaches. Most qualitative approaches stress
emotions and the self, information-control about self collecting copious amounts of data before delving
and illness, meanings of chronic illness, and ways of into the analysis; researchers using such approaches
living with it. often complete their major analytic work long
Different chronic illnesses give rise to variations after they have left the field. In contrast, grounded
and complexity in conceptual development of the theorists use their emerging theoretical categories to
crosscutting themes. For example, some chronic ill- shape the data collection while in the field as well as
nesses resulted in periodic, progressive, or permanent to structure the analytic processes of coding, memo-
visible disabilities; other disabilities remained invisi- making, integrating and writing the developing
ble. The kinds of situations posed by such differences theory [20-231. The ‘groundedness’ of this approach
were studied as well as how ill people thought and fundamentally results from these researchers’ com-
felt about them, and what effects visible disability mitment to analyze what they actually observe in the
made when studying relations between self, emotions, field or in their data. If they find recurrent themes or
information, and time. issues in the data, then they need to follow up on
My social constructionist version of using the them, which can, and often, does lead grounded
grounded theory method to study illness makes use of theorists in unanticipated directions. For example,
four different phases in developing concepts and while a graduate student, I conducted a study of
theoretical frameworks: (1) creating and refining the caring for ill and dying elders in working-class
research and data collection questions, (2) raising families [24]. An unexpected theme emerged about
terms of concepts, (3) asking more conceptual ques- the role coroner’s deputies played in notifying
tions on a generic level, and (4) making further families about the death of their relative. Checking
discoveries and clarifying concepts through writing out that theme led me to further comparative re-
and rewriting. In this paper, I shall illustrate each search with coroner’s deputies in different settings
phase and will draw my examples a@ discussion of about their strategies for announcing death [25].
substantive work mainly from two recent papers, Grounded theorists begin with general research
‘Disclosing Illness’ [ 171,an examination of ill people’s questions rather than tightly framed pre-conceived
constructions about what, when, and how to tell hypotheses. If, perchance, those research questions
others about their illness, and ‘Struggling for a Self: are irrelevant in the field, then they develop new,
Identity Levels of the Chronically III’ [I 81, an analysis suitable ones or find another field. For example, my
of how and why ill people construct identity hier- prior experiences as an occupational therapist and
archies with specific levels. Throughout my analysis as a researcher in rehabilitation led me to expect
below, I stress the actice stance and approach of the to find chronically ill people suffering gradual or
researcher since I see being active as crucial to using rapid decline. However, suffering and decline varied
the grounded theory method [19]. enormously despite routine clinical descriptions of
specific illnesses as downhill and physicians’ dire
THE DISCOVERY PERSPECTIVE predictions for certain patients. Hence, the research
questions were revised accordingly to tap ill people’s
The grounded theory method lived experiences and their constructions of them.
Before applying the grounded theory method to Throughout the research and writing processes,
chronic illness, a brief outline of its defining charac- grounded theorists follow interests, leads, and
teristics will bring it into clearer focus [20-231. Like hunches that they find or identify in the data. Then
other analytic approaches, the grounded theory they may gather more data, ask more questions, and
method itself offers a way of constructing sociological check their developing categories. Their emergent
reality; using the method fosters developing analytic categories explain and conceptualize (1) the data,
and conceptual constructions of the data. In their (2) common sense understandings of these data,
sociological constructions. grounded theorists aim to and, likely, (3) other theoretical interpretations. For
create theoretical categories from the data and then example, some people seem to ‘deny’ illness. But
analyze relationships between key categories. In when viewed from an ill person’s vantagepoint of
short, the researcher constructs theory from the desiring to realize identity goals and struggling to
data. By starting with data from the lived experience have a valued self, that person’s behavior becomes
of the research participants, the reasearcher can, understandable, rather than standing as evidence of
from the beginning, attend to how they construct ‘denial of illness’ [18]. To illustrate, an ill woman
their worlds. That lived experience shapes the continues to work while having a serious episode of
‘Discovering’ chronic illness 1163

illness. She does so because she wishes to avoid the analysis. Generally, much qualitative research
having illness inundate her life and her identity, not depends on implicit methods and thus, relies on
because she disavows its presence. the researcher’s intuition and talent. In contrast,
With grounded theory strategies, theoretical devel- grounded theory specificies an explicit set of analytic
opment turns on theoretical sampling [20-231. Here, guidelines and procedures which can help any quali-
the researcher collects new data to check, fill out, and tative researcher develop more or less fruitful concep-
extend theoretical categories. In contrast to Strauss tualizations of his or her data.
[23], I conduct theoretical sampling only after I have
defined key concepts. Delaying focused theoretical
sampling fosters gaining an in-depth understanding Debates and dilemmas
of the realities and issues at hand. Hence, theoretical Clearly, grounded theory differs from survey re-
sampling fits into the research and analytic process search and from other types of qualitative research
much later than initial sampling of sites, people, or such as ethnomethodology. Variations also exist
documents. By the time theoretical sampling is among proponents of the grounded theory method
planned, a researcher would have some hunches or and within the same proponent at different points in
even hypotheses which he or she wishes to check. time. My social constructionist perspective represents
Thus, theoretical sampling shapes further data collec- one variation. Glaser and Strauss’ [22] early work
tion as the researcher pursues developing conceptual represents another. They (221 developed grounded
ideas rather than amassing general information. For theory when formal theory was becoming more arid
example, my research led to gathering more materials and distant from the worlds of interacting people,
to specify and clarify the conditions under which ill when the primary value of qualitative research lay in
people form identity goals and to delineate conditions sharpening later ‘rigorous’ quantitative research, and
when they revise their goals. when field research studies generally remained ethno-
When developing formal theory and forming graphic or consisted of analytic description. They
more generic concepts, grounded theorists take their attempted to demonstrate that the strategies of
substantive analysis from one area and conduct grounded theory: (1) brought the researcher close to
theoretical sampling in other substantive areas. Then, basic processes and issues that people experience;
they develop and refine their original analysis by hence, theoretical constructs at once revealed and
using constant comparative methods [22]. For covered those realities; (2) provided a method for
example, my concept of an identity hierarchy [18] identifying, capturing and rendering processuaI
reflects ill people’s rankings of their identity goals rather than static analyses; (3) fostered developing
after experiencing loss and uncertainty. To develop a rigorous qualitative methodology with its own
the concept, identity hierarchy, for a formal theory, integrity and intrinsic values distinct from quanti-
one could check the concept first with people who tative research; and (4) offered possibilities from
also face issues of unexpected change, loss, and moving qualitative research more definitively toward
uncertainty. Thus, a grounded theorist could apply dense, durable, substantive and formal theories [27].
the concept to people who with little warning: (1) lose Misunderstandings and criticisms. Like all other
seemingly secure jobs without having ready access to methods in social research, the grounded theory
equivalent positions and (2) must immigrate such as method has strengths and weaknesses. A number of
certain Central Americans or Southeast Asians. By the criticisms of grounded theory reflect an incom-
studying these people, the grounded theorists might plete understanding of the logic and strategies of the
discover other contributing conditions which influ- method.
ence the creation of an identity hierarchy. Then data Such partial understanding can lead to applying
from contrasting sources could be sampled to help inappropriate criteria on which to judge the method
develop variation [26]. Moving across substantive [28-301. For example, Bulmer [28, p. 6671 states that
areas fosters developing conceptual power, depth, Glaser and Strauss’s [I S] “tabula rasa view of inquiry
and comprehensiveness. is open to serious doubt.” He accuses them of
The strategies of grounded theory can yield rich espousing pure induction since they propose reading
data, elaborated categories, and dense analyses with the literature in the relevant fields after having devel-
applications across substantive fields. These strategies oped a set of categories.
do not yield statistically verified results, which require Instead, reading and integrating the literature later
random sampling of a clearly identified population. in the research process is a strategy to prompt
However, survey researchers may test reported exploring various ways of analyzing the data. But
grounded theories through standard logical-deduc- it means only delaying the literature review, not
tive verification procedures. overlooking it, or failing to use it. Delaying the
Grounded theory studies aim for analytic power literature review decreases the likelihood that the
and conceptual grasp which synthesize, explain, researcher will already be locked into preconceived
and interpret the data. The rigor of the grounded conceptual blinders upon entering the field and in
theory method depends upon developing the range of interpreting the data. Once the researcher has devel-
relevant conceptual categories, saturating (i.e. filling, oped a fresh set of categories, he or she can compare
supporting, and providing repeated evidence for) them with concepts in the literature and can begin to
those categories, and explaining the data. Similar place his or her study appropriately within it. As
to quantitative studies, or any other research, the Glaser [21] stresses, grounded theorists must do their
quality of grounded theory studies varies according own analytic work; if they ‘borrow’ concepts from the
to the methodological thoroughness, the significance literature, then they should ensure that these concepts
of the research questions, and the incisiveness of merit a place in their analysis.
II64 KATHY CHARMAZ

Although Glaser and Strauss are not explicit in The direct, and vivid words. My earlier jargon included
Discorery of Grounded Theory, they do assume that “mobilizing the self” 1331, and “encapsulated time”
researchers have had solid training in their fields. [33], which I have since dropped, and in these two
Also, they assume that such training provides re- more recent pieces, “supernormal social identity” [ 181
searchers with perspectives from which to observe and “cathartic spilling” [ 17). Several more vivid terms
and on which to build analyses, not merely to apply include “the restored self” [18], “merged identities”
them. Glaser and Strauss’ method fosters both using [ 18, 331, “making a comeback” [33] and “the salvaged
disciplinary and theoretical perspectives and con- self” [18].
tributing to their development with new ideas, dense In addition to the weaknesses in using the method,
analyses, and theory construction. Emerson [29] and problems also may arise from lack of clarity about
Katz [30] reveal another misinterpretation. Emerson key terms in grounded theory. Then, such questions
(29, p.971 states “while grounded theory glorifies and arise as: “When has a researcher achieved saturation
tries to further generate theory in its own right, it also of a category?,” or “What stands as a theory?” One
treats discovery as a stage prior to verification. This researcher’s conceptual framework may resemble
rigid divorce between discovery and verification lends another researcher’s theory [35]. Qualitative ‘theoriz-
support to the critique of fieldwork as insightful but ing’ often remains discursive and imbedded in de-
not rigorous.” First, Glaser and Strauss [22] portray scription, which may reduce it to a loosely integrated
the contrast between discovery and verification ap- conceptual framework. In my view, a theory expli-
proaches starkly because they see the need for devel- cates a phenomena, specifies concepts which catego-
oping new lines of theoretical development in the rize the relevant phenomena, explains relationships
discipline. Second, qualitative research generally, and between concepts and provides a framework for
grounded theory specifically, derive from different making predictions.
canons than logico-deductive verification models. Phenomenological and positicistic emphases in
Qualitative research typically stresses inductive, grounded theory. The grounded theory method of
open-ended, intuitive approaches to data-gathering analysis is just that, a method. My social construc-
and analysis, particularly in the early stages of both. tionist version of grounded theory has a phenomeno-
Third, precisely what grounded theory provides is logical cast [36]. Glaser and Strauss’ earlier works
a rigorous method for qualitative studies. Hence, [2 I, 221 have both phenomenological and positivistic
grounded theory must be assessed from internal emphases and therefore, sometimes may seem confus-
logic of its own method, not by the inappropriate ing and even inconsistent. They claim to be phe-
application of external criteria founded in other nomenological, yet a strong positivistic thread runs
methods [3 11. through their work. On the phenomenological side,
How researchers actually use the method remains Glaser and Strauss have always emphasized going
a different issue than whether the method itself directly to the real world and starting with issues in
possesses rigor and logical consistency. Here, Katz it. They then went into hospitals and studied expec-
and Emerson come close to reifying traditional verifi- tations of death from detailed first-hand field obser-
cation methods as the only scientific method and vations and interviews [37].
therefore, as the only legitimate scientific work. On the positivistic side, the early works on the
Most criticisms of grounded theory turn on mis- grounded theory method suggest that the method
understandings or misuse of the method. However, takes on a life of its own, independent of its propo-
the major problems with the grounded theory method nents and independent of the researcher. Glaser and
lie in glossing over its epistemological assumptions Strauss [22, p. 341 state, “Our approach, allowing
and in minimizing its relation to extant sociological substantive concepts and hypotheses to emerge first,
theory. The relation between subjectivist and objec- on their own, enables the analyst to ascertain which,
tivist realities and levels of explanation remains un- if any, existing formal theory may help him generate
specified. And ways in which grounded theorists use his substantive theories.” Similarly, Glaser [21, p. 51
their prior theoretical perspectives remain somewhat reiterates, “Grounded theory arrives at relevance,
ambiguous. because it allows core problems and processes to
Potential weaknesses in using the method. Weak- emerge. ” Also, Glaser implies that core processes and
nesses in using the method may have become equated problems will similarly reveal themselves, rather than
with weaknesses inherent in the method. Such weak- be defined by actors and analysts. Hence, Glaser and
nesses may also be found in most other types of Strauss lean toward assuming that the theoretical
qualitative research and in quantitative research as categories derive from the data and that the re-
well. These weaknesses include: premature commit- searcher remains passive [21,22]. Here, they come
ment to a set of analytic categories [29,30], unneces- close to positing an external reality, unaltered by
sary jargon, and a lack of clarity about key terms the observer’s presence. Whether they intended to
such as theory, category, and saturation. Premature do so or simply had a theoretical lapse in the midst
commitment to categories means that the researcher of methodological claims-making is itself open to
has not fully explored the issues, events, and mean- construction. Clearly, however, Strauss’ [23] recent
ings within the research problem or setting and has explication of grounded theory now reveals an
not gained what Lofland call “intimate familiarity” actively involved researcher who constructs cate-
with it [32]. gories and concepts. Yet, as Denzin [38] observes,
Like other sociological perspectives, the grounded Strauss’ [23] grounded theory approach remains with
theory method does lend itself to generating unneces- the empirical science tradition and displays the ten-
sary. esoteric jargon for labeling categories. Instead, sion between being simultaneously subjectivist and
researchers can label their categories with simple, scientific.
‘Discovering’ chronic illness 1165

A SOCIAL CONSTRL’CTIONIST GROtiNDED THEORY ness. One could then look at how ill people invent
liveable worlds.
A social constructionist grounded theory views the Similarly, having had considerable experience with
process of categorization as dialectical and active, chronically ill people may foster developing certain
rather than as given in the reality and passively lines of analysis, depending on the nature of that
observed by any trained observer. Hence, a social experience. For example, Lubkin’s [42] nursing text
constructionist perspective assumes an active, not (which does not take a social constructionist view)
neutral, observer whose decisions shape both process stresses both patient compliance and professional
and product throughout the research (391. In short, advocacy. A social constructionist could take those
the research report is also a social construction of the topics and analyze how professionals construct and
social constructions found and explicated in the data. act upon their definitions of and criteria for patient
The interaction between the researcher and the data compliance and professional advocacy.
result in ‘discovering’, i.e. creating, categories. In The researcher’s values-in the case of chronic
short, the ‘discovery’ process consists of the re- illness, toward aging, disability and death-may
searcher creating discoveries about the data and shape the research process in taken-for-granted ways.
constructing the analysis. How the analyst uses the Further, an over-riding interest in one concept such
method and which questions he or she brings to the as stigma or, for that matter, in one specific chronic
data shape the results. Certainly having a store of illness, fundamentally shapes the later analysis [43].
sociological concepts cues a researcher to look for Perhaps most crucial is the researcher’s school of
constructions around status, power, hierarchy, and thought, which provides the conceptual roots for the
self-concept [40]. In addition, the researcher’s per- categories to grow. Following a social constructionist
spective leads to asking certain kinds of questions perspective fosters creating categories of the research
[41]. In my studies, they include: How do chronically participants’ beliefs and actions. In order to create
ill people construct definitions of their illness, recov- categories, the researcher already must have a firm
ery, seif and situation? Who benefits from their grounding in sociological concepts without being
definitions? How do their definitions compare with wedded to them. This stance implies a delicate
larger cultural themes and social values? What do balance between possessing a grounding in the disci-
these people take as real in their situations; of pline and pushing it further. In my work, concepts
what are they aware; unaware? How do they feel such as identity, stigma, awareness, and meaning
about others, themselves, their plans, hopes and inform both the data collection and the analysis.
prospects? Under which conditions do ill people have When wedded to concepts in their disciplines,
relative autonomy and control? Under which do they researchers may neither see beyond them nor use
not? them in new ways. For example, a medical sociologist
From a social constructionist view, the researcher who uses stigma as a definitive concept [16] to order
takes those questions a step further. Whether ad- and integrate data, rather than as a sensitizing con-
dressing definitions, awareness, feeling, control, or cept [16], may only apply it as given in the literature
any experience, the social constructionist attempts instead of using it to ask new questions and to form
to find how each develops, changes and gives rise new leads. Sensitizing concepts [16] alert researchers
to consequences. For example, which do ill people to central issues to tap without committing them to
believe contributes to their having control over their reproducing the initial set of concepts.
lives? How do they define that control? In which ways Using the grounded theory method, on the one
do their conceptions of control reflect larger ideo- hand, necessitates developing, refining, revising, and
logies? How do their constructions of control develop transcending concepts within the discipline. Often, a
and change over time and over the course of their social constructionist stance elicits a fresh look at
illness? What consequences emerge from their con- existing concepts. That alone contributes to revising
structions of control? and refining them. On the other hand, using
Whatever their particular philosophical stance, grounded theory means dealing with the rendering
grounded theorists build in special data-gathering the actual research data (i.e. interviews, documents,
questions. based on their assumptions and substan- case histories, accounts etc.). When social construc-
tive interests. For example, given my substantive tionists study their data, they continually raise the
interests in the social psychology of time and in the questions: “How?“; “Why?“; “Under which condi-
sociology of emotions, I often ask questions such as: tions?“; “With which consequences?” How do people
“What was time like for you during your hospitaliz- construct beliefs? How do they manage their iices? Why
ation?’ “ How did you feel about telling your parents do they think, feel, and act the way that they do? Under
about your diagnosis?” “Tell me about what time was which conditions do they think, feel, and act that way?
like when things were so uncertain.” What are the consequences of their beliefs, feelings,
The researcher’s perspective consists of more than and actions? The ‘grounded’ nature of this research
philosophical stance, school of thought, and method- strategy is three-fold: (1) researchers attend closely to
ological strategies. It also consists of experiences, the data (which amounts to ‘discoveries’ for them
values, and priorities. Upon reading my data, a when they study new topics or arenas), (2) their
colleague interested in the family sees how family theoretical analyses build directly on their interpreta-
interactions affect communications and decisions tions of processes within those data, and (3) they
about illness. One could take that perspective and must ultimately compare their analyses with the
examine how families construct definitions of illness extant literature and theory.
and decisions about the ill person. Another colleague In short, the flexibility inherent in field work more
interested in science sees these ill people’s inventive- generally assumes special importance in grounded
1166 KATHY CHARMAZ

theory. Grounded theorists start with what they see DEVELOPING AND REFINING RESEARCH AND
happening sociologically, and then they interpret it. DATA COLLECTION QUESl-lONS

What happens in the setting or to the research


participants shapes the collected materials. For exam- The research questions
ple, studying chronic patients’ motivations for being By asking how, why and under which conditions an
productive did not interest me, despite practitioners’ existing sociological concept works in this specific
urging me to pursue the topic. Yet over and over field, social constructionists can use existing concepts
again, ill people tried to explain their hopes, plans as sensitizing concepts as Blumer [!6] urges. Such
and intentions-in short, motivations. Hence, the sensitizing concepts become translated into general
data drew me to address this topic. research questions and into more concrete data col-
Again, the researcher’s perspective influences what- lection questions. How pointed these questions are
ever he or she sees within the data. Certainly Mead’s depends on the type and level of theoretical develop-
[9] theory of emergence, analysis of action, and ment that the researcher has reached.
perspective on the development of self and Strauss’ Generally, grounded theorists start with a set of
[lo] explication of identity shaped my perspective. experiences they wish to explore-in this case,
Since my study focused on issues around self-concept chronic illness and how it affects the self. From the
and identity, some questions were asked around those beginning of my doctoral research, I wished to ex-
topics and respondents volunteered much other infor- plore links between having chronic illness, experien-
mation. Their responses indicated that their motiva- cing time, and shaping the self. Hence, these research
tions vitally concerned them. In addition, their questions initiate the inquiry: How do chronically ill
motivations were intertwined with their identity people view living with chronic illness? In which ways
goals. Without an initial theoretical interest in iden- does having a serious chronic illness affect an individ-
tity, one might miss that connection. ual’s shaping of self? How is the shaping of self
These interpretations of the data dovetailed with related to the individual’s experience of time? After
my theoretical interests in ways which spurred con- completing one study [33] on those topics, another
ceptual development. My questions reflected social study was initiated later with 70 more interviews of
constructionist concerns and often rapidly built on ill people and caregivers and a repertoire of published
my developing concepts. This strategy led to shaping and unpublished accounts about or by people with
data collection around sociological and analytical chronic illness.
interests rather than, in this case, gathering quasi- Exploring the thematic questions led me to asking
medical information, or making strictly behavioral a set of interview questions that tapped people’s
recordings. From the beginning of the research, the chronologies of illness and how it impinged upon
grounded theory approach fosters the researcher their lives. The chronologies gave some sense of the
staying on an analytic path, albeit he or she may range of experiences people had had as well as how
identify a range of possible analytic paths (see Fig. 1). they cast them. Given my interests, questions about

Save for a later


research project

MEMoWRmNG

later research

MEMowRrruuG
to concepts

OFEARUER
DATA

Fig. 1. Research levels in the grounded theory method.


‘Discovering’ chronic illness 1167

identity and self were built in from the beginning. others were involved. “Tell me what happened when
From those questions and the spontaneous remarks you had the crisis” allows the researcher to piece
ill people made about themselves, codes and cate- events and people together before asking about help-
gories were developed describing, synthesizing and ful participants. Both loaded questions above may
explaining the data. prove to be useful, when suitable and when raised in
logical sequence, i.e. if the researcher senses that the
Deceloping interview questions patient played a part in the decision or already knows
To use the grounded theory method effectively, the that other people were involved in the crisis. If not,
researcher needs rich, detailed data. Grounded theo- both questions may raise sticky issues at a time the
rists have been accused of not attending carefully to respondent may not be able to face them.
data collection and of skimping on sampling [32]. Briefly, interview questions can be framed and
However, grounded theorists need detailed, vivid ordered by developing these kinds of questions and
data on which to base their analyses. Unlike some leads: (1) short face-sheet, (2) informational, (3)
qualitative researchers, grounded theorists may show reflective, (4) feeling, and (5) ending. The short
less concern with ascertaining the ‘accuracy’ of a face-sheet questions are intended to be neutral, fac-
specific piece of data than with its theoretical rele- tual and limited to necessary information. These
vance. Two people, such as an ill person and his or questions set the tone for the interview, so researchers
her caregiver, may give different account for the ill need to be aware of their tone and mode of asking
person’s activities. Here, whose rendering of reality them.
comes closer to ‘truth’ has less importance than the The informational questions bring the respondent
analytic issues raised within each view, as well as the further into the interview and establish chronology,
conflicting definitions of each participant. The possi- types of events, degrees of awareness, cast of partici-
bility of someone’s account being inaccurate raises pants, and the like. If a researcher has established
interesting theoretical considerations for constructing rapport, he or she can bring in reflective and feeling
an identity. questions. When trust and ease come more slowly,
In order to develop a durable, useful grounded then sequencing them carefully works better.
theory, however, the data must provide a variety of Given my theoretical interests, the reflective ques-
complete accounts of major issues and processes. By tions may often serve as transitions to address direct
obtaining these data, the grounded theorist can then issues about self: “How did - affect you?’ “How
depict their research participants’ worlds. Such data did you see yourself then?’ “How would you
results in developing analyses more easily with more compare the person you were - years ago with the
convincing arguments. person you are now?” Similarly, feeling questions
Unlike most other types of qualitative data, also often directly elicit data about self. “Tell
grounded theorists explicitly shape the materials they me what you were feeling when you learned the
gather. Learning to shape the materials well from the diagnosis.” “ How did you feel about taking the early
start provides the basis for later coding and categoriz- retirement?” All of these questions help to elicit the
ing. Further, interviewing works well in studies of narrative of the respondent’s story with only minimal
chronic illness since the researcher often wants to framing by the researcher [46]. Ending questions are
obtain detailed individual chronicles. In addition, designed to complete the interview on a positive note
participant observation may be impossible for certain [44]. The more intense the interview, the more ques-
types of research problems in this area. tions and comments needed to end the interview with
Framing, pacing, and managing the interview ques- the person feeling positive about self. These questions
tions all affect the type and quality of material the also elicit insights about self. “What have you learned
researcher obtains [32,44,45]. Since my studies relied about yourself over the past - years?’ “How have
heavily on interview data, the incisiveness of the you grown as a person after having these experi-
analysis turned on developing suitable questions and ences?’ Such questions elicit interesting data on
knowing when to ask them. Given the nature of symbolic meanings of illness and self as well as
chronic illness, several interviews may be needed with accomplishing positive closure.
a respondent just to get through the basic infor-
mation about the course of his or her illness, much Coding and categorizing the data
less tap all the areas the researcher needs to cover. To the extent possible, the codes and categories
As Lofland and Lofland [32] put it, the interview reflect emerging ideas rather than merely describing
should be a “directed conversation”. How to direct topics. For example, I coded some observations of
the conversation depends on the respondent’s present elderly married couples as “merged identities”
psychological and physical status, the relationship [l&33] rather than as family relationships, a more
between researcher and respondent, the researcher’s general topic. The codes and categories then help the
theoretical perspective, and the topic. Being too researcher begin to take the data apart and frame
directive poses hazards when interviewing chronically analytic questions about it. In that way, the codes
ill people. The overly-directive researcher can cut off and categories help the researcher to build an analysis
the most interesting leads and rich data. Further, the of the data rather than remain at the level of
researcher may load assumptions into the questions ethnographic description.
without being aware of doing so [44]. Asking “How By grounding the categories in the data as speci-
did you decide to have the surgery?” assumes that the fically yet analytically as possible, the grounded
respondent decided. In contrast, “How did you come theorist can then sharpen the category, suggest
to have the surgery?’ leaves things open. “Who was its parameters, begin to outline the conditions
most helpful to you during the crisis?’ assumes that under which the category develops, and start to look
1168 KATHY CHARMAZ

for consequences of it. Conditions mean those pre- closing or avoiding disclosure. (Data coded in one
requisites which influence and shape views, inter- way earlier can be coded several other ways. Simi-
actions, and events, rather than determine them. larly, the same anecdote may point to several differ-
By maintaining this type of analytic posture, even ent conceptual issues.) Subsequently, I built questions
the initial ways of handling the data can enhance about disclosing into later interviews. Doing so
developing a theoretical analysis out of it. helped me to frame a more complete picture of it.
A few suggestions for coding. Line by line coding The level of skill of the researcher in collecting data
during the initial coding prompts the researcher to enters here too. As a doctoral student, I knew little
study the data, to dispel1 earlier preconceived as- about interviewing and initially, felt uncomfortable in
sumptions about the data, and to begin viewing the doing it. As a result, the quality of the early data
data analytically [20-221. As a grounded theorist differed strikingly with that of the later interviews. In
begins to render some codes into categories, he or she the interim, I had gained both skills and confidence.
defines them analytically and delineates their proper- Since then, I have discovered that many researchers
ties. However implicitly. the researcher’s definitions are neither particularly skilled at interviewing nor do
of reality clearly shape what categories he or she they design artful questions. Even practiced inter-
constructs. Line by line coding keeps the researcher viewers may lapse into ‘do you’ questions which elicit
examining the colIecred data, rather than lapsing ‘yes’ or ‘no’ answers with limited elaboration.
entirely into theoretical flights of fancy which have
little connection to the data. Yet the researcher can RAISING TERMS TO CONCEPTS
invoke his or her theoretical perspective to raise
questions about the data. My theoretical interests, for Raising terms to concepts means that the re-
example, lead to such questions as: (1) Of what larger searcher takes a term or code, defines it succinctly,
process is this action a part? (2) How did this action and analyzes it. The wording of the term is important
(belief, definition, relationship, pattern or structure) since the researcher now intends to treat it as a
evolve? (3) What do these data state or assume about conceptual category, rather than merely as a descrip-
self and about relationships? By examining the tice topic, or code. In contradistinction to quanti-
collected data with the theoretical eye, the set of tative studies, here a code means a label applied to
categories developed remains closer to the actual data certain data; a code remains less abstract than a
and simultaneously, moves beyond description. conceptual category. In turn, a conceptual category is
Coding for processes, actions, assumptions, and part of the researcher’s larger theoretical framework
consequences rather than for topics leads to greater in which he or she specifies conditions, offers explana-
analytic precision. Coding for processes assists in tions, and makes predictions. When treating a term
defining major activities and issues; such coding also as a conceptual category, the researcher specifies its
helps the researcher to discern connections between properties, notes the conditions under which it arises,
structures and events. It helps to look at topics relates how it changes, describes its consequences,
processually. Thus, rather than treating my code, and ultimately, specifies its relationship to the other
supernormal identity [18] only as a topic, I treated it conceptual categories. Descriptive topics, in contrast,
more processually by looking at the development and may provide vivid accounts but are not integrated
abandonment of supernormal identities. Doing so into a set of specified and inter-related concepts
fosters studying the conditions which gave rise to from which the researcher develops explanations or
these identities. Similarly, coding for actions foster2 predictions. For example, I referred to each rung of
the researcher’s search for processes. The researcher the identity hierarchy as an identity level [18] and
can then look for phases. contributing conditions and categorized each level. Then, I specified the con-
consequences of those actions. ditions under which someone moved up or down the
A major strength of the grounded theory method identity hierarchy in order to realize a preferred
is its open-endedness and flexibility. Since analysis identity. In that way, a set of related conceptual
and data collection proceed simultaneously, a re- categories were developed and subsequently, the
searcher can follow up on ideas as he or she creates relations between them could be developed.
them. Thus. researchers may use the method to Two analytic processes contribute to raising terms
examine an issue thoroughly rather than to rely on to concepts-constant comparison and continued
previously collected data, which may address the questioning. The comparisons include data with data,
issue incompletely. A grounded theorist may sculpt category with category and concept with concept
fully contoured ideas throughout the analysis rather [20-231. In my study, comparing data with data
than only suggesting or alluding to them at the end. means: (I) comparing different people’s situations,
The open-endedness of grounded theory allows beliefs, behavior or accounts of the same type of
researchers to pursue leads and ideas as they develop. event or issue, (2) comparing data from the same
For example. most of the people who wished to lead people at different times and (3) comparing properties
conventional lives recounted incidents concerning found in the data with other properties. Comparing
either disclosing illness to others or avoiding dis- different people’s accounts means taking some topic
closure entirely. Whatever they did, issues around such as an experience, issue, period of time, relation-
telling people about their illness loomed large to ship, or stage and juxtaposing data from each person
them. Hence. I went back through earlier interviews against each other one. For example, comparisons
and sorted material on disclosing and avoiding dis- were made of ill people’s experiences of discovering
closure. Looking through the interviews with this and defining illness. Then accounts were compiled to
interest now in mind caused me to find numerous compare the circumstances under which they discov-
comments and accounts of incidents concerning dis- ered and defined illness, how they felt, thought and
‘Discovering’ chronic illness 1169

dealt with it, what they faced at that time, who researcher moves directly into analysis of the data.
became involved, and so forth. Just looking at such Bits of data and early codes are systematically exam-
basic issues systematically yields thick description ined, explored and elaborated upon. Through memo-
and often, analytic insights. On a more abstract level, writing, the researcher takes his or her emerging ideas
comparing identity goals as illness changed became apart, checks them, and outlines further data collec-
another unit of comparison [ 181.Thus, I learned that tion. During each stage of memo-writing, the re-
some people do plummet down the identity hierarchy searcher may use his or her theoretical background to
as they experience their illnesses. Moreover, a few deepen the analytic insights of his or her developing
others start at the bottom of the hierarchy and over grounded theory. My interest in constructions of
months or years move up in it. reality, for example, alerted me to compare people’s
Raising a term to a conceptual level means making accounts of disclosing or avoiding disclosure and
a series of decisions about it. Again, the researcher thereby, to assess the relative amount of reality that
actively shapes the research process. The researcher they granted to illness.
creates an explication, organization, and presentation Basically, memo-writing gives the researcher a too]
of the data rather than discovering order within the for engaging in an extended on-going dialog with self
data. The discovery process consists of discovering [47,48]. By committing ideas, hunches, questions,
the ideas the researcher has about the data after and elaborated categories to organized memos, the
interacting with it. Raising a term to a conceptual researcher defines what is implicit and what is explicit
level first means deciding that the term reflects a in the data. In that dialog with self, the researcher
significant process, relationship, event, or issue. looks at the data from a variety of perspectives and
Second, it means explicitly deciding to follow up on analyzes them. Later, the researcher may impose
it in subsequent data collection. Third, it means organization upon the memos by ordering them to
making connections between it and other conceptual best capture the data and communicate with the
categories. audience. In the meantime, memo-writing gives the
Again, such terms are not always ones a researcher researcher an analytic handle on the materials and a
expects to study. For example, I did not intend to means of struggling with discovering and defining
study disclosing illness nor did I have any particular hidden or taken-for granted processes and assump-
interest in that topic until issues around difficulties in tions within the data. For example, my initial respon-
disclosing or avoiding disclosure appeared in many dents’ statements led me to sense that the concept of
interviews. Ill people attached such significance to identity encompassed more for them than social
these issues, which, in turn, caused me to look at them location or personal definition. But what was it?
more systematically and to raise new questions about As I gathered more data, I realized that they
them. formed identity goals, which took a hierarchical form
To raise the term disclosing to a conceptual level, [18,33].
I looked first at my data and then to the literature in The dialog with self through memo-making typi-
medical sociology. I first asked basic questions such cally helps to separate the researcher from the re-
as: searched, thereby reducing problems of immersion in
the setting or data, ‘going native’, assuming the
(1) What did disclosing mean to ill people?
stance of the practitioner and the like. The finished
(2) When were they concerned about it?
memos form a repository of ideas, which the re-
(3) What were the consequences of it?
searcher can then rethink, revise, toss out, organize,
(4) How did they handle disclosing?
and present in varied ways. In a larger sense, memo-
(5) When did ill people view disclosing as irrele-
writing may never have closure since a researcher
vant?
may make continued theoretical discoveries as his or
For many of these people, disclosing meant reveal- her memos become increasingly conceptual. The later
ing something private, often potentially discrediting, memos then may include, incorporate, and transcend
about themselves. The link between disclosing and ill extant theory as the researcher develops his or her
people’s emotions was clear. Later, how these issues grounded theory.
intertwined with self-images and self-concept became
evident. Therefore, I looked at how people managed WRITING AND REWRITING
disclosing. This meant studying how ill people con-
structed their actions of disclosing. I raised questions The discovery process also extends into writing and
about other ways of telling: were there other types of rewriting drafts for publication. The researcher gains
telling and, if so, how did they compare and contrast further insights and creates more ideas about the data
with disclosing? My awareness of modes of telling while writing. Hence, writing and rewriting actually
was heightened by an earlier study of coroner’s become crucial phases of the analytic process.
deputies announcing death to survivors [25]. I had Through writing and rewriting, a researcher can
long realized that physicians and other practitioners identify arguments and problems, make assumptions
announce, inform, impart, or even reveal bad news. explicit, and sharpen the concepts [48]. Further, the
But they seldom disclose for disclosures reveal some- writing process gives the researcher the opportunity
thing about self. to link his or her work with other theories by
integrating them into the discussion and analysis.
Memo writing At this point in the analytic process, the researcher
Memo-writing provides the pivotal step of break- may have a theoretical grasp of the material. But he
ing the categories into components and elaborating or she may not have formed the analysis into an
the codes [20-231. Through memo-writing, the argument or presented it as a problem of interest to
1170 KATHY CHARMAZ

colleagues or practitioners for publication. Now, the them, make links to the literature, critically examine
researcher needs to do a thorough review of the the categories and concepts, and present the data
literature in the field. Then he or she can frame the cogently [Jj]. In writing the article cited above,
study within that literature and show where and how renaming three of my categories enhanced clarity and
it fits. The following excerpts show how I located my vividness. Hence, the ‘reconstituted self’ (it sounded
study on identity and framed the problem [18, p. 2871. like processed food) became the ‘restored self’, which
People with serious chronic illnesses struggle to have captured the idea of returning to a past self and
valued lives and selves. Those who live with serious chronic reclaiming former identities. The ‘possible personal
illness experience disrupted plans and altered lives. Their identity’ became the ‘contingent personal identity’
illnesses may cause setbacks, flare-ups, complications, im- since the possibility remained contingent upon no
paired functions and disabilities as well as have social. further episodes or complications. A ‘preferable self-
psychological, and financial consequences. All these experi- view’ became the ‘salvaged self’ because ill people did
ences and consequences of illness affect whether ill persons more than offer a self-image they preferred; they tried
lead valued lives and realize their hopes to live on their own
to salvage prior defining traits to claim a valued self
terms. Their illnesses pose identity problems that often are
left entirely to them and their significant others. How do
now.
they handle these identity problems? I answer this question Writing and rewriting certainly fosters analytic
by analyzing the experiences of chronically ill people whose clarity. These processes also prompt gaining more
former identities and future plans become questioned, un- theoretical comprehensiveness and precision as the
dermined. altered. or negated. By addressing their elusive researcher grapples with increasingly more abstract
struggle for a self, my analysis focuses on the consequences theoretical questions and hones his or her responses
of experiencing illness for creating self and identity rather to them.
than concentrating primarily on the practical struggles of
managing life with chronic illness.
SU~IMARV AND DlSCUSSION
This paragraph sets the stage for studying the
problems of struggling for a self and creating valued Explicating use of a social constructionist version
identities. The category of people who have these of grounded theory method to study chronic illness
problems are specified-those with serious chronic reveals issues concerning the method more generally
illnesses whose former identities have become ques- as well as offering some specific guidelines for socio-
tioned. In addition, the focus of the analysis is logical research in chronic illness. Rather than reflect-
pinpointed, as contrasted with another major theme ing a tubulu rasa, grounded theorists bring to their
in studies of chronic illness-managing to live with it. studies the general perspectives of their disciplines,
In that article, I turn next to the theoretical analysis their own philosophical, theoretical, substantive, and
of identity and introduce the major ideas and argu- methodological proclivities, their particular research
ments. Hence, the reader can discern the framework interests, and their biographies. They do not bring,
of the analysis and how the pieces fit together. however, a set of finely-honed preconceived concepts
Key ideas are as follows [I 1, pp. 283-2851: and categories to apply automatically. Should
Chronically ill persons’ accounts of their experiences grounded theorists apply such concepts and cate-
reveal two new interconnected issues for studying identity: gories-even their own previous ones-to new data,
the role of preferred identities and the development of iden- they must justify them.
tity hierarchies. First, individuals choose identities by out- When using the grounded theory method, re-
lining plans and assessing prospects. The chronically ill people searchers actively form questions and seek data. The
whom I interviewed implicitly developed preferred identities experiences of people with chronic illness do not
as they attempted to construct lives apart from illness. Their entirely unfold before researchers’ eyes. Rather, re-
preferred identities symbolize assumptions, hopes, desires searchers create their analyses. The questions that
and plans for a future now unrealized. In short preferred researchers put to the world, how they collect their
identities mean identity goals.. . the concept of identity
encompasses the person’s vision of future selves, reflecting
data, and which isssues and processes they see within
his or her hopes, aspirations, objectives and goals. In this it all fundamentally shape their analyses. Further,
sense, preferred identities serve as a source of motivation. researchers create a conceptual interpretation of the
Second, an identity hierarchy becomes visible as ill data, impose an order on it, explicate the relation-
people, over time, choose different types of preferred iden- ships between categories, and organize those relation-
tities, reflecting relative difficulty in achieving specific aspi- ships to communicate their ideas to audiences. Thus,
rations and objectives. The types of preferred identities like quantitative researchers [50], grounded theorists
constitute particular identity levels in the identity hierarchy. make decisions throughout the research process that
These identity levels include: (1) the supernormal social shape their research products. Also like quantitative
identity, an identity demanding extraordinary achievement
in conventional worlds; (2) the restored self, a reconstruc-
researchers. grounded theorists’ research reports may
tion of previous identities before illness; (3) contingent take an objectified form and tone that belies the
personal identity, a hypothetically possible, though uncer- actual research process.
tain, identity, because of further illness; and (4) the salvaged Using the grounded theory method to study
self, retaining a past identity based on a valued activity or chronic illness offers the researcher strategies for
attribute while becoming physically dependent. Experi- focusing and controlling large amounts of data in
encing progressive illness often means reducing identity ways that render it conceptually and, in turn, move
goals and aiming for a lower level in the identity hierarchy. the emergent conceptualizations toward more general
In short reducing identity goals means aiming for a less theoretical statements. Such statements not only
preferred identity.
deepen sociological understanding of the experience
Through writing and rewriting, the researcher can of chronic illness, but also, contribute more generally
bring out implicit arguments, provide a context for to the discipline.
‘Discovering’ chronic illness 1171

The grounded theory method provides the possibil- discover the conditions which promote this theoreti-
ity of theory development. The analyses flowing from cal sophistication instead of theoretical tunnel-vision
this approach are ones with conceptual power and may, in itself, require grounded theory research.
durability. But, like other analyses, they can always
be refined and modified.
Acknowledgemenrs-Thanks are due to Adele Clarke, Uta
Since grounded theorists believe in studying pro- Gerhardt, Anna Hazan, Marilyn Little and Barbara Rosen-
cess, they realize that theories cannot be frozen in blum for reviewing an earlier version of this paper and to
time. Certainly changing historical conditions can Uta Gerhardt and Anselm Strauss for their comments on
alter any area of inquiry. Grounded theory analyses the preceding draft. I also wish to thank Adele Clarke for
can be adapted to changing conditions and can either rewording the title of the paper.
take historical change into account-or even focus
upon it. Grounded theory analyses can also provide
future researchers with sources of conceptual, REFERENCES
explanatory, and predictive comparison.
From my perspective, grounded theory analyses 1. Strauss A. L. et al. Chronic Illness and the Quality of
Life. Mosby, St Louis, MO, 1984.
can be enriched by clarifying the researcher’s episte-
2. Fagerhaugh S. Getting around with emphysema. In
mological premises and by reaching back into extant Chronic Illness and the Quality of Life (Edited by Strauss
theory. To date, grounded theorists have not expli- A.). pp. 99-107. Mosby, St Louis, MO, 1975.
cated a shared set of epistemological premises. Per- 3. Reif L. Ulcerative colitis: strategies for managing life. In
haps there should not be such a set of premises, for Chronic Illness and the Quality of Life (Edited by Strauss
researchers from varied backgrounds with diverse A.), pp. 81-88. Mosby, St Louis, MO, 1975.
research problems can use the strategies of grounded 4. Weiner C. L. The burden of rheumatoid arthritis. In
theory. If so, then individual researchers need to Chronic Illness and the Quality of Life (Edited by Strauss
examine their own epistemological premises as well as A.), pp. 71-80. Mosby. St Louis, MO, 1975.
5. Charmaz K. Loss of self: a fundamental form of
those imbedded in their use of the grounded theory
suffering in the chronically ill. Social. Hlth f//n. 5,
method. Doing so could clarify the relation between 169-195, 1983.
the subjective and objective, sharpen the research 6. Schutz A. Collected Papers. Volume I: The Problem of
process, and delineate the theory of reality to which Social Realify. Martinus Nijhoff, The Hague, 1971:
the researcher subscribes. Currently, most grounded Berger P. and Luckmann T. The Social Construction of
theory pieces, including my own, assume a theory of Reality. Doubleday, New York, 1966.
reality. Similarly, some pieces seem to assume the 7. Foucauh M., The Birth of the Clinic. Pantheon, New
discovery of objectified truth, rather than the discov- York, 1973.
ery of the researcher’s more or less useful theoretical 8. Lindesmith A., Strauss A. and Denzin N. Social PSI-
chology. Prentice-Hall, Englewood, Cliffs, NJ, 1988.
categories.
9. Mead G. H. Mind, Self and Society. University of
Researching back into extant theory might seem to Chicago, Chicago, IL, 1934.
contradict the entire grounded theory approach. I 10. Strauss A. L. Mirrors and Masks. Sociology Press. Mill
think not. As noted throughout the above analysis, a Valley, CA, 1969.
strong theoretical perspective can guide questions- II. Weigert A. J., Teitge J. and Teitge D. Society and
from the basic research outline directly into the data Identity. Cambridge University Press, Cambridge, 1986.
gathering and analysis. In that way, a strong theoret- 12. Denzin N. On Understanding Emotion. Jossey-Bass. San
ical perspective fosters giving the research greater Francisco, CA, 1984.
conceptual depth and breadth while firmly situating it 13. Merleau-Ponty M. Phenomenology of Perception. Rout-
ledge, Kegan Paul, London, 1962.
within the discipline. More direct ties back to their
14. Berger P. and Luckmann T. See Ref. [6].
Meadian heritage could, for example, sharpen 15. For a counter-argument positing the dismissal of struc-
grounded theorists’ theory of action and assumptions tural issues by social constructionists, see Gerhardt U.
about human nature implicit in their studies. Further, Ideas about Illness: An Intellectual and Political History
a greater attention to contemporary developments of Medical Sociology. Macmillan, London. In press.
in Marxist and critical theory might foster closer 16. Blumer H. Symbolic Interactionism. Prentice-Hall,
connections between microscopic and macroscopic Englewood Cliffs, NJ, 1969.
structures in grounded theory analyses. Such em- 17. Charmaz K. Disclosing illness. Unpublished manu-
phases would neither contradict nor negate prior script, 1986.
work, but would supplement and extend earlier 18. Charmaz K. Struggling for a self: identity levels of the
chronically ill. In The Experience and Management of
ground-breaking efforts. Chronic Illness. Research in the Sociology of Health Care
Nonetheless, bringing a strong theoretical perspec- (Edited by Conrad P. and Roth J.), Vol. 6, pp. 283-307.
tive to grounded theory studies does pose some JAI Press, Greenwich, CT, 1987.
knotty problems. Prior theoretical socialization in a 19. A grounded theorist makes decisions, follows leads,
researcher may produce ideational and ideological organizes data, defines categories, develops concepts
baggage which inhibits forming fresh ideas and pro- and demonstrates the relations between them. A caveat.
motes tunnel-vision. Yet, theoretical sophistication in Researchers in other fields may profitably use the
classical and contemporary sociological theory can grounded theory method. I stick to sociological con-
cerns as they reflect my interests and expertise.
also foster asking fundamental questions throughout
20. Charmaz K. The grounded theory method: an explica-
the research and analytic processes. Such sophistica- tion and interpretation. In Contemporary Field Research
tion leads the researcher beyond training for theoret- (Edited by Emerson R. M.), pp. 109-127. Little Brown,
ical sensitivity to create sound analytic questions. Boston, MA, 1983.
Rather, such sophistication leads the researcher to 21. Glaser 8. G. Theoretical Sensithity. Sociology Press,
address basic questions of value and meaning. To Mill Valley, CA, 1978.
1172 KATHY CHARMAZ

22. For the first statement of the method. see Glaser B. G. M. R. Social constructionism and the development
and Strauss A. L. The Discovery of Grounded Theory. of medical sociology. Social. Hllh Illn. 8, 137-167.
Aldine. Chicago, IL. 1967. 1986.
23. A more explicit formulation of steps and techniques 37. See Glaser B. G. and Strauss A. L. Awareness of Dying.
clarifies and greatly extends the Discovery book. See Aldine, Chicago, IL, 1967; and Time for Dying. Aldine,
Strauss A. L. Qualilatice Analysis for Social Scientists. Chicago, IL, 1968.
Cambridge University Press, New York, 1987. In keep- 38. Denzin N. Review of Strauss A. L. Qualirafit,e Methods
ing with the grounded theory method, I did not read the for Social Scientists. In Conremp. Social. 17, 430-432,
then just-published volume of Qualitarice Analysis until 1988.
after I had written two drafts of this paper. 39. For a critique of value-free sociology, see Gouldner A.
24. Calkins K. Shouldering a burden. Omega 3, 23-36, The Coming Crisis of Western Sociology. Avon, New
1972. York, 197f.
25. Charmaz K. The coroners’ strategies for announcing 40. Gerhardt U. The Parsonian paradigm and the identity
death. Urbcrn L* 4, 296-316, 1975. of medical sociology. Social. Ret). 27, 229-251, 1979.
26. For a discussion of theoretical sampling, see Ref. [23, 41. Such questions hint of ways that grounded theory may
pp. 276-2771. be useful in looking at social psychological issues in a
27. Bigus O., Hadden S. C. and Glaser B. G. Basic social context which addresses macro structural issues. For
processes. In Qualiralic~e Methods: Volume II of Hand- a further discussion of analyzing stressful life events in
book of Social Science Methods (Edited by Smith R. B. a theoretical context which takes macro issues into
and Manning P. K.), pp. 251-272. Ballinger, Cam- account. see Gerhardt U. Coping and social action:
bridge, MA, 1982. theoretical reconstruction of the life-event approach.
28. Bulmer M. Concepts in the analysis of qualitative data. Social. Hlth Illn. 1, 195-225, 1979.
Social. Rer. 27, 651-677, 1979. 42. Lubkin I. M. Chronic Illness: Impact and Intementions.
29. Emerson R. M. (Ed.) Introduction, Part II: theory and Jones & Bartlett, Boston, MA, 1986.
evidence in field research. In Contemporary Field 43. Locker D. Disability and Disaduanrage: The Conse-
Reseurch, pp. 93-108. Little Brown, Boston, MA, 1983. quences of Chronic Illness. Tavistock, London, 1983.
30. Katz J. A theory of qualitative methodology. In Con- 44. Charmaz K. Intensive interviewing. Unpublished
temporary Field Research (Edited by Emerson R. M.), manuscript, 1986.
pp. 127-148. Little Brown, Boston, MA, 1983. 45. Gordon R. L. Interoiewing. Dorsey Press, Homewood,
31. For another discussion of the application of inappro- IL. 1980.
priate criteria, see Star S. L. Triangulating clinical and 46. Mishler E. G. Research Inrerciewing: Conrexf and
basic research: British localizationists, 1870-1906. Hisr. Narrafire. Harvard University Press, Cambridge, MA,
Sci. 24, 29-48, 1986. 1986.
32. Lofland J. and Lofland L. H. Analyzing Social Settings. 47. Memo-writing comes close to what teachers of writing
Wadsworth, Belmont, CA, 1983. call pre-writing or free-writing, although memoes are
33. Charmaz K. Time and Identity: The Shaping of Selves of focused on a category or code. See, for example, Elbow
the ChronicalIy Ill. PhD dissertation, University of P. Writing with Power. Oxford University Press, New
California, San Francisco. CA, 1973. York, 1981. The analyst may write anything in the
34. I am indebted to Susan Leigh Star for the term. memo without the constraints of evaluation, or of
35. See Clarke A. A social worlds adventure: the case of audiences. Memo-writing helps to reduce writer’s block
reproductive science. In Theories of Science in Society and also helps to bring the fluidity, imagery, and rhythm
(Edited by Cozzins S. and Gieryn 1.). Indiana Univer- of spoken language into the work.
sity Press, Bloomfield, IN. In press. Given my definition 48. Becker H. S. Writing for Social Scienrisrs. University of
of theory, most grounded theory analyses actually Chicago Press, Chicago, IL, 1986.
provide incisive conceptual frameworks, which aim 49. For a way of presenting data on interviews and respon-
toward theory, rather than qualifying as theory. dent accounts which specifies the amount, if any, of
36. My perspective comes close to phenomenology in that rendering by the researcher, see Gerhardt I-l. and
I also emphasize lived experience. It departs from Brisekorn-Zinke M. The normalization of hemodialysis
phenomenology in that I do not pursue a quest for at home. In Research in the Sociology of Health
essences as phenomenologists do. Adopting a social Care (Edited by Roth J. A. and Ruzek S.), Vol. 4,
constructionist perspective leads to a certain reflexivity pp. 271-317. JAI Press, Greenwich, CT 1986.
in examining one’s research methods. For a detailed 50. Fletcher C. Beneath the Surface: An Account of Three
examination of social constructionism in medical sociol- Styles of Sociological Research. Routledge & Kegan
ogy with an emphasis on medical knowledge, see Bury Paul, London, 1974.