Embedded Case Study Methods

Embedded Case Study Methods
Contributors: Roland W. Scholz & Olaf Tietje
Editors: Roland W. Scholz & Olaf Tietje
Book Title: Embedded Case Study Methods
Pub. Date: 2002
Access Date: October 15, 2013
Publishing Company: SAGE Publications, Inc.
City: Thousand Oaks
Print ISBN: 9780761919452
Online ISBN: 9781412984027
DOI: http://dx.doi.org/10.4135/9781412984027.n2
Print pages: 9-15
This PDF has been generated from SAGE Research Methods. Please note that the
pagination of the online version will vary from the pagination of the print book.
Copyright ©2013
SAGE Research Methods
[p. 9 ↓ ]
The case study approach presented is an empirical inquiry that investigates a
contemporary problem within its real-life context. Understanding the problem and its
solution requires integrating a myriad of mutually dependent variables or pieces of
evidence that are likely to be gathered at least partially by personal observation.
Although a common definition of case studies exists, one may encounter various types
of case studies (see Table 2.1). In order to make clear to which type of case study the
introduced methods of knowledge integration should be applied, we will briefly describe
different types of case studies. A detailed review of case studies is given by Yin (1994).
Holistic Versus Embedded
A crucial distinction must be made between holistic and embedded case studies (Yin,
1994, p. 41). A holistic case study is shaped by a thoroughly qualitative approach that
relies on narrative, phenomenological descriptions. Themes and hypotheses may be
important but should remain subordinate to the understanding of the case (Stake, 1976,
p. 8).
Embedded case studies involve more than one unit, or object, of analysis and usually
are not limited to qualitative analysis alone. The multiplicity[p. 10 ↓ ] of evidence is
investigated at least partly in subunits, which focus on different salient aspects of the
case. In an organizational case study, for example, the main unit may be a company as
a whole, and the smallest units may be departments or even groups of individuals, such
as owners and employees. In a clinical, neuropsychological case study, the units may
be organized along biographically critical events in the childhood or the vocational world
Page 2 of 7
Embedded Case Study Methods: TYPES OF CASE
Copyright ©2013
SAGE Research Methods
of the case. In case studies on regional or urban planning, the units may be different
interest groups that are involved or affected by the project.
Table 2.1 Dimensions and Classifications of Case Studies
Note that an embedded case study allows for a multiplicity of methods that may be
applied within the subunits. Thus, hypotheses may be formulated, quantitative data
sampled, or statistical analyses applied (see Bortz & Döring, 1995; Campbell & Stanley,
1963). As the title of the book suggests, this book presents methods of embedded case
studies (see Part III).
Single Case Versus Multiple Case
Another design characteristic of a case study is whether the design is single case or
multiple case. There may be different reasons for choosing a[p. 11 ↓ ] single-case
design. A case may be considered unique, prototypical, salient, or revelatory to the
understanding of a phenomenon or problem. Analogous to Newton's experimentum
crucis, it may even be the critical case in testing a well-formulated theory. Although
there is no common understanding of how to integrate separate single-case studies
into a joint multiple-case design, it is most important to note that the synthesis process
between the single cases does not follow a statistical sampling rationale. As Yin (1994)
notes, “Every case should serve a specific purpose within the overall scope of inquiry.
Here, a major insight is to consider multiple cases as one would consider multiple
experiments—that is, to follow a ‘replication’ logic” (p. 45).
Page 3 of 7
Embedded Case Study Methods: TYPES OF CASE
Copyright ©2013
SAGE Research Methods
The case study researcher often feels intrinsically motivated to investigate a certain
case for nonscientific reasons. This may hold true for a new type of educational or
public health program, or a specific project in urban development. If there is intrinsic
interest, the study team usually takes responsibility and is accountable for the analysis
and its consequences (see Gibbons et al., 1994). But if the objective of the study
is something other than understanding the particular case, then the inquiry is an
instrumental case study.
To illustrate the difference between these types of studies, consider the characters of
two different physicians. A physician with an intrinsic motivation is personally interested
in and feels responsible for the patient. A physician with an instrumental motivation
is primarily interested in using anamnestic and laboratory data to further scientific or
financial objectives, and is less interested in the case itself.
The label case study is most frequently associated with the exploratory case study. It
usually precedes a final study, which can, itself, be a case study, but it can also have
a different research design (Boos, 1992). Exploratory case studies help to gain insight
into the structure of a phenomenon in order to develop hypotheses, models, or theories.
An exploratory study very much resembles a pilot study; the research design and data
collection methods usually are not specified in advance.
[p. 12 ↓ ]
A descriptive case study differs from an exploratory study in that it uses a reference
theory or model that directs data collection and case description. In some respects, a
descriptive case study tests whether and in what way a case may be described when
approaching it from a certain perspective. Many Formative Scenario Analyses may be
considered typical of this type of study (see Chapter 9).
Page 4 of 7
Embedded Case Study Methods: TYPES OF CASE
Copyright ©2013
SAGE Research Methods
Explanatory case studies can also serve to test cause-and-effect relationships. Clearly,
according to conventional understanding of theory testing, a single case can only falsify
a theory. However, a case may also be used for theory testing, either if the case is
used for quantitative data sampling (see Petermann, 1989), or, in a replication logic,
if the research team investigates “whether similar causal events—within each case—
produce these positive outcomes” (Yoon & Hwang, 1995, p. 12). Note that the theory
testing is done in a qualitative manner. However, as in traditional hypothesis testing,
specifications for the cause-impact chain have to be formulated before case analysis.
A case study may be used as a method of research, teaching, or action/ application.
For instructional purposes, case studies are commonly used in business, law, and
medical schools. The case encounter quite often changes the traditional educational
approach into a discussion pedagogy. Thus, the case method is a variation on the
Socratic method, which is another name for proactive interaction between teachers and
students (Ronstadt, 1993). Unfortunately, when teaching by case studies (see Barnes
et al., 1994), the primacy of data and of situation analysis is often not respected as a
principle. This is due to the fact that a prepared, written case offers only limited access
to data, and, therefore, teaching case studies are based on a virtual process of case
Several basic formats for case studies exist (Ronstadt, 1993, pp. 17–18). The first two
types are teaching cases and are always provided in written form.
[p. 13 ↓ ]
Page 5 of 7
Embedded Case Study Methods: TYPES OF CASE
Copyright ©2013
SAGE Research Methods
In principle, each case study should use multiple sources of information. All methods
should employ direct and participant observations, structured interviews, and surveys,
and they can also include experimental design, focused interviews, open-ended
interviews, archival records, documents, and scientific data from field and laboratory
(see Box 2.1). (A detailed description of data gathering is given in Yin, 1994, p. 93, and
Stake, 1995, p. 49.) This remains true regardless of case design. The main distinction
for case studies is whether they have a holistic or embedded design.
Knowledge integration within a holistic design is ruled almost exclusively by the
principles of qualitative research. The synthesis process is[p. 14 ↓ ] informal (avoiding
reductionism and elementalism), empathic, and mostly intuitive. Thus, the research
report is narrative in nature.
Box 2.1 Using Multiple Sources of Data and
In all phases of the case, a wide variety of data from different sources have to be
integrated (Yin, 1994, p. 91). The source and type of data depend on the case and its
Documents, archival records, and open-ended interviews are typical sources used in
the beginning of most studies. In an embedded case design, structured or focused
interviews are often used, but this design also allows for surveys, questionnaires, and
even the sampling experimental data. In neuropsychological or environmental case
studies, laboratory data or simulation studies are also helpful in gaining insight into the
case. The following figure illustrates the potential sources of evidence and techniques
for data sampling that can be integrated in case analysis.
Page 6 of 7
Embedded Case Study Methods: TYPES OF CASE
Copyright ©2013
SAGE Research Methods
The embedded case design allows for both qualitative and quantitative data and
strategies of synthesis or knowledge integration. The methods provided in Part III of this
volume may be used to interrelate and integrate the variables, findings, evaluations,
and soon from the various facets of the case or subunits of case inquiry. Thus, the
methods of knowledge integration help explain the data under consideration, thereby
making data and inferential processes more transparent. The global statements and
conclusions are usually derived by an intuitively qualitative process based on both
experiential understanding and a more or less formative synthesis process that is
supported by the methods introduced.
Page 7 of 7
Embedded Case Study Methods: TYPES OF CASE