Tải bản đầy đủ
3: The Use of Interview Data

3: The Use of Interview Data

Tải bản đầy đủ


Case Studies 173

Figure 10.1 Developing Grounded Theory Through the Case Study Method
Research Begins

Grounded Theory

• Offer explanations for
the problem or issue
originally considered
• Compare your
explanations with those
current in the literature

Research Idea
• Consider some
problem or
issue for study


• Review and consider
your research
• Consider what the
findings mean
• Assess the implications for
the conceptual framework
• Seek disconfirming evidence
• Seek alternative
• Compare findings with
the literature


• Organize and code
• Place concepts into
conceptual frameworks
• Link data to research
themes and the

research question might be rather general, and the interview subjects would be selected from a pool of potentially
interchangeable members of the study population. In the
case study interview, the topic under study can be quite
focused, and each individual included in the study might
be chosen due to their particular connection to the case.
In one recent example, Susan Sullivan and Vivian
Shulman set out to examine a period of significant managed change in a New York City school district. The
changes were planned in response to new school mandates, and the researchers initially collected data from all
of the different schools in the district, as well as the central office. “When it became clear that the leadership of
the superintendent drove all district practices, we shifted
our focus to describing his particular style and its efficacy
in promoting systemic change” (Sullivan & Shulman,
2005, p. 124). The study was still about the school district
changes, but the focus shifted from studying a case of
negotiation among offices to a case of leadership under


• Review literature
• Determine single
or multiple cases
• Establish access plan
• Identify data-collection
• Consider analytic

Data Collection

• Use data-collection
plan from design
• Recognize the need
to be flexible
• Consider literature;
begin comparisons
and analysis as
data accrues

constraint in a time of transition. The principal datacollection method was one-on-one interviews with the
superintendent, his staff, teachers, and principals in the
affected schools.
The advantage of the case study over the individual
interview is demonstrated in this example. Sullivan and
Shulman spoke at length with the superintendent about
his education and leadership philosophies. But they were
able to triangulate their data by interviewing others
about the ways in which the superintendent attempted to
implement his philosophies. The additional data from the
district schools provided a real-life measure of the impact
of the district leadership on those who were responsible
for running the individual schools. Further, rather than
speaking with teachers or administrators about how
school districts operate, the researchers focused on the
specific teachers and administrators in that one district,
and asked questions that were also particular to that
one setting.

174 Chapter 10


Throughout the preceding paragraphs, the chief suggestions for information (data) gathering have been the
use of interviews and observation. As implied earlier,
however, all forms of qualitative research may constitute a case study. You should, therefore, be familiar with
the possible use of records concerning the subject. For
individuals, these may include birth, marriage, divorce,
property ownership, and educational records of the subject. For groups and associations, there may be meeting
minutes, documents of incorporation, and official publications. For either, you might include an assortment
of other more or less official documents such as police
actions, court records, evaluations, and so forth. All of
these official documents are potentially valuable sources
of information in a case study. As well, an individual
subject might have written personal documents, letters,
diaries, or blogs. Both individual and collective subjects
are likely to have created a trove of e-mails that can be
included in the study.

10.3.1: The Use of Personal Documents
The general use of personal documents is discussed
in Chapter 8. As suggested there, personal documents
involve any written record created by a person that concerned his or her experiences. The common types of documents classified under this label include autobiographies,
diaries and journals, letters, and memos written by a
subject in a research investigation. In addition, and given
the extent to which people use photographic and video
equipment today, these items may also serve as categories
of personal documents.
Autobiographical documents include a considerable
variety of written material. They may be published or
unpublished documents, cover an entire life span, or focus
on only a specific period in a subject’s life or even a single
event. Even a written confession to a crime may be seen by
some researchers as a type of autobiographical document.
Certain manifestos may serve as both personal statements
and confessions, though this is uncommon.
Diaries and journals also may arise in a number of
varieties. A diary may be kept with no purpose in mind
beyond the writer’s personal desire to maintain a record
of daily events. It may be maintained in order to provide
some therapeutic release or as a kind of log and chronological listing of daily events during new experiences, such
as an internship. Or, a diary or journal may be created at
the specific request of a researcher as a contribution to
some study. In the latter case, one may consider the material in a solicited document (see Chapter 8).
Letters provide an intriguing view into the life of
the author. Typically, letters are not created by the writer
with the intention of having them used by a researcher.

As a result, they frequently reflect the inner worlds of
the writer. They may record the writer’s views, values,
attitudes, and beliefs about a wide variety of subjects. Or,
they may describe the writer’s deepest thoughts about
some specific event or situation about which they report.
Historians have long seen the value of letters to document
events during the past. Letters written by military figures
and politicians, for example, may allow researchers to
better understand how and why certain battles have been
fought. Letters written by criminals such as serial killers
and bombers provide insight into how the culprit thinks
and potential explanations for their actions. Letters are
simply replete with potentially useful information.
The use of memoranda, including e-mails, has become
commonplace in virtually all organizational settings.
Memos may contain strictly work-related information or
casual insider jokes and communications. They may reflect
the tone and atmosphere of a work setting, as well as the
potential level of anxiety, stress, and morale of the writers.
Moreover, they may even show the research aspects of
the workplace culture or work folkways. Also, they may
contain information relevant to understanding the general
organizational communications network used in the setting, the leadership hierarchy, various roles present in the
setting, and other structural elements. Thus, a memorandum can provide an interesting self-disclosing aspect of
its creator, or various aspects of a group or organization,
when used as data in a case study.
Photographic and video equipment has become so
commonplace, and oversharing so common, that many
people now regularly record and publish their lives
and the lives of their family members in this manner. It
becomes important, therefore, for researchers to consider
how these items may illustrate various aspects of the
subject’s life and relationships. This may involve stepping
back and examining an entire photograph in terms of what
it shows in general; it may include an examination of the
expressions of people shown in a picture; it could involve
consideration of where a picture or video was taken or
recorded such as on a vacation, in the home, or at a party;
or it may involve determination of the reason the photograph or video was created—as a simple family record to
commemorate some situation, to have as a keepsake, to
document some event or situation, to brag, to entertain,
and so forth.
The literal value of personal documents as research
data is frequently underestimated in contemporary
research texts and courses. Although such documents
are certainly extremely subjective in their nature, this
data should not be viewed as a negative or, in this case,
even as some sort of limitation or shortcoming. It is the
very fact that these documents do reflect the subjective
views and perceptions of their creators that makes them

useful as data in a case study. It is precisely through this
subjectivity that these documents provide information
and insight about the subject that might not be captured
through some other more pedestrian data-collection

10.4: Intrinsic, Instrumental,
and Collective Case
10.4 Classify three types of case studies
Stake (1994, 1995) suggests that case studies can be usefully
classified into three different types: intrinsic, instrumental,
and collective.
Intrinsic case studies are undertaken when a researcher
wants to better understand a particular case. It is not
undertaken primarily because it represents other cases or
because it illustrates some particular trait, characteristic, or
problem. Rather, it is because of its uniqueness or ordinariness that a case becomes interesting (Creswell, 1998, 2007;
Stake, 1994, 2000). The role of the researcher is not to understand or test abstract theory or to develop new or grounded
theoretical explanations; instead, the intention is to better
understand intrinsic aspects of the particular child, patient,
criminal, group, organization, event, or whatever the case
may be (Munhall, 2007). The case may generate findings
or stimulate ideas that will be applicable to other cases, but
the intrinsic case study is not performed for those reasons.
Erikson’s case study of the Buffalo Creek flood started off
that way. The study needed to be done because something
important had happened there.
Instrumental case studies provide insights into an issue
or refine a theoretical explanation, making it more generalizable (Creswell, 2002; Stake, 1994). In an instrumental
case study, the researcher focuses on a single issue or concern and identifies a single case to illustrate this item of
focus or concern (Creswell, 2007). In these situations, the
case actually becomes of secondary importance, playing
a supportive role (Denzin & Lincoln, 2005). The details of
the case provide a background against which the larger
research interests will play out. Instrumental case studies
are often investigated in depth, and all aspects and activities are detailed but not simply to elaborate the case per
se. Instead, the intention is to help the researcher better
understand some external theoretical question, issue, or
problem. Instrumental case studies may or may not be
viewed as typical of other cases. However, the choice of a
particular case for study is made because the investigator
believes that his or her understanding about some other
research interest will be advanced.

Case Studies 175

Stake (1994, 2000) also points out that because researchers often have multiple interests, there is no solid line drawn
between intrinsic and instrumental studies. In fact, a kind
of “zone of combined purpose separates them” (Stake, 1994,
p. 237). My study of the origins and development of the field
of organized collective action in response to HIV/AIDS in
New York City (Lune, 2007) began in that combined zone.
I viewed the AIDS example as an important case of collective action, but it took a while to define what factors defined
this case or made it comparable to other cases.
Collective case studies (Stake, 1994, 2000, 2005) are
also known as multiple-case studies, cross-case studies,
comparative case studies, and contrasting case studies
(Gerring, 2006; Merriam, 2001). Collective case studies
involve extensive study of several instrumental cases,
intended to allow better understanding, insight, or perhaps improved ability to theorize about a broader context. Yin (2003a) argues that multiple cases may be
selected in order to try replicating insights found in
individual cases or to represent contrasting situations.
Regardless of one’s purpose, Yin (2003a, p. 46) indicates
that multiple-case studies are frequently “considered
more compelling, and the overall study is therefore
regarded as more robust.”
In each of these three approaches, the validity of the
research hinges on how we address the question, “What
is this a case of?” Researchers interested in important but
uncommon events might well choose a comparative case
study approach in order to gather relatively large amounts
of related data on their topic despite its lack of frequency.
Revolutions, for example, do not occur very often. A single
revolution can be rooted in a long and complex local history, and supported by myriad factors that can only be
found through a deep understanding of the global economy and political alignments. Yet, by comparing the French
Revolution of 1789 with the American Revolution of 1776 and
the Cuban Revolution of 1959, researchers can identify crucial
similarities and patterns that greatly help us to understand
revolutionary events. But, not so fast. Are these cases best
understood as cases of “revolution,” or would it be more
useful to call them cases of “successful revolutions”? After
all, there have been many failed revolutions as well, which
share many of the same precursors as the successful ones.
Does it make sense to include revolutions across generations,
or should we focus on cases of twentieth-century revolutions distinct from eighteenth- or nineteenth-century ones?
Researchers also need to decide on other defining characteristics for their cases. Are revolutions that overthrow autocracies
and install democracies significantly different from revolutions that go in the other direction? Are those that overthrow
democracies better characterized as coups? All of these challenges point to the importance of the conceptualization stage
of research design, as discussed in Chapter 2.

176 Chapter 10


10.5: Case Study Design
10.5 Classify four types of case study designs
According to Yin (1994, 2003a) and Winston (1997), there
are several appropriate designs for case studies: exploratory, explanatory, and descriptive. These categories, or
variations on them, are often used to distinguish among
the different orientations to any research. It is worthwhile
thinking about how each of these orientations plays out for
case study research.

10.5.1: Exploratory Case Studies
When conducting exploratory case studies, fieldwork
and data collection may be undertaken before defining
a research question. This type of study may be seen as a
prelude to a large social scientific study—which may or
may not in itself involve case studies. From our perspective, the study must have some type of organizational
framework that has been designed prior to beginning the
research. Others, however, such as Yin (2003b, p. 6), suggest that these exploratory case studies may follow “intuitive paths often perceived by others as sloppy.” But, as
Yin (2003b) also points out, the goal in these studies may
be justified when they seek to discover theory through
directly observing some social phenomenon in its natural
and raw form. The sort of exploratory study may be useful
as a pilot study, for example, when planning a larger, more
comprehensive investigation (Swanson & Holton, 2005).
Calling a study “exploratory” should not be an excuse
for failing to plan. The circumstances under which such an
approach would be valid include needing to respond quickly
to unanticipated events and, as suggested above, exploring a
topic or setting in order to design the follow-up research.

10.5.2: Explanatory Case Studies
Explanatory case studies are useful when conducting causal
studies or otherwise pursuing an inferential research question. Particularly in complex studies of organizations or
communities, one might desire to employ multivariate cases
to examine a plurality of influences. The explanatory case
study, then, attempts to discover and analyze the many
factors and conditions that can help us to build a causal
explanation for the case. We do this for theory development,
theory testing, and theory expansion. Theory development
case studies are useful when something new or unexpected
has occurred, and we can only begin to build a theoretical
model to explain it by comparison to other cases that might
be said to resemble our case. We know that some of these
other cases can help explain the new case, but we don’t yet
understand the uniqueness of the present case.

10.5.3: Descriptive Case Studies
With descriptive case explorations, the investigator presents
a descriptive theory that establishes the overall framework
for the investigator to follow throughout the study. What
is implied by this approach is the formation and identification of a viable theoretical orientation before enunciating
research questions (Hancock & Algozzine, 2006; Munhall,
2007). The investigator must also determine before beginning the research what exactly the unit of analysis in the
study will be. For example, if I define a particular disaster
as an “industrial” disaster and not a “natural” disaster or
an accident, then I am claiming that the kinds of factors that
tend to cause industrial disasters are better explanations
for the current case than the factors usually associated with
other types of failures or related problems. My case study
would be designed to measure the factors that I believe are
most relevant, and also to measure the factors that I think
are less relevant. I can then compare the influence of the
different measures to see how well my interpretation holds
against the other possibilities. Descriptive case studies differ
from explanatory ones in that we would be focused on the
uniqueness of the case and not try to develop an inferential
model that would necessarily be applied to other cases.
Jason Jensen and Robert Rodgers (2001, pp. 237–239)
offer a different typology of case studies. They recommend
distinguishing among “snapshot” case studies that occur
at one point in time, longitudinal studies conducted over
a fixed period of time, and “pre- and post-”event studies. They also identify comparative case studies in which
one focuses on the significant differences between two or
more otherwise comparable cases. Their typology also recognizes that case studies may use any combination of the
above. That is to say, just because we can distinguish types
of studies among different categories does not mean that
you have to commit to just one of those categories.

10.5.4: Designing Case Studies
Designing a case study is merely a special case of the problem of designing any study, as discussed in Chapter 2. All
studies begin with a research question, or problem to be
addressed. Most are designed around testable propositions
derived from theory and the existing research literature.
Whatever data-collection methodologies we adopt, all of
the primary data must concern the same unit of analysis.
In this context, the unit of analysis defines what the case
study is focusing on (what the case is), such as individuals,
a group, an organization, a city, and so forth. If we were
studying the response to an epidemic, for example, then
all of our data collection would be organized in terms of
conceptual variables that define that response, though our
background sources would certainly address other things
such as population dynamics, disease etiology, and the
social, economic, and cultural context of our case.