A case study uses
as many data sources as possible to systematically investigate individuals, groups,
organizations, or events. Case studies are conducted when a researcher needs to
understand or explain a phenomenon. They are frequently used in medicine, anthropology,
clinical psychology, management science, and history. Sigmund Freud wrote case
studies of his patients; economists wrote case studies of the cable TV industry
for the FCC; and the list goes on and on.
Case study
research includes both single cases and multiple cases. Comparative case study
research, frequently used in political science, is an example of the multiple case
study technique.
Merriam (1988)
lists four essential characteristics of case study research:
1. Particularistic. This means that the case study focuses on a
particular situation, event, program, or phenomenon, making it a good method for
studying practical, real-life problems.
2. Descriptive. The final product of a case study is a detailed
description of the topic under study.
3. Heuristic. A case study helps people to understand what’s being studied. New
interpretations, new perspectives, new meaning, and fresh insights are all
goals of a case study.
4. Inductive. Most case studies depend on inductive reasoning. Principles and
generalizations emerge from an examination of the data. Many case studies
attempt to discover new relationships rather than verify existing hypotheses.
Conducting a Case Study
Five distinct stages in carrying out a case
study: design, pilot study, data collection, data analysis, and report writing.
- Design. The first concern in case study design is what to ask. The case study is most appropriate for questions that begin with “how” or “why.” A research question that is clear and precise focuses the remainder of the efforts in a case study. A second design concern is what to analyze. What constitutes a “case”? In many instances, a case is an individual, several individuals, or an event or events. If information is gathered about each relevant individual, the results are reported in the single or multiple case study format
- Pilot Study. Before the pilot study is conducted, the case study researcher must construct a study protocol. This document describes the procedures to be used in the study and also includes the data-gathering instrument or instruments. A good case study protocol contains the procedures necessary for gaining access to a particular person or organization and the methods for accessing records. It also contains the schedule for data collection and addresses logistical problems. For example, the protocol should note whether a copy machine is available in the field to duplicate records, whether office space is available to the researchers, and what supplies are needed. The protocol should also list the questions central to the inquiry and the possible sources of information to be tapped in answering these questions. If interviews are to be used in the case study, the protocol should specify the questions to be asked. Once the protocol has been developed, the researcher is ready to begin the pilot study. A pilot study is used to refine both the research design and the field procedures. Variables that were not foreseen during the design phase can emerge during the pilot study, and problems with the protocol or with study logistics can also be uncovered. The pilot study also allows the researchers to try different data-gathering approaches and to observe different activities from several trial perspectives. The results of the pilot study are used to revise and polish the study protocol.
- Data Collection. At least four sources of data can be used in case studies.Documents, which represent a rich data source, may take the form of letters, memos, minutes, agendas, historical records, brochures, pamphlets, posters, and so on. A second source is the interview. Some case studies make use of survey research methods and ask respondents to fill out questionnaires; others may use intensive interviewing. Observation/participation is the third data collection technique. The general comments made about this technique earlier in this chapter apply to the case study method as well. The fourth source of evidence used in case studies is the physical artefact—a tool, a piece of furniture, or even a computer printout. Although artifacts are commonly used as a data source in anthropology and history, they are seldom used in mass media case study research. (They are, however, frequently used in legal research concerning the media.) Most case study researchers recommend using multiple sources of data, thus permitting triangulation of the phenomenon under study (Rubin, 1984). In addition, multiple sources help the case study researcher improve the reliability and validity of the study.
- Data Analysis. Unlike quantitative research techniques, there are no specific formulas or “cookbook” techniques to guide the researcher in analyzing the data. Consequently, this stage is probably the most difficult in the case study method. Although it is impossible to generalize to all case study situations, Yin (2003) suggests three broad analytic strategies: pattern matching, explanation building, and time series.
In the pattern-matching strategy, an
empirically based pattern is compared with one or more predicted patterns. For
instance, suppose a newspaper is about to initiate a new management tool:
regular meetings between top management and reporters, excluding editors. Based
on organizational theory, a researcher might predict certain outcomes—namely,
more stress between editors and reporters, increased productivity, and weakened
supervisory links. If analysis of the case study data indicates that these results
do in fact occur, some conclusions about the management change can be made. If
the predicted pattern does not match the actual one, the initial study propositions
have to be questioned.
In the analytic
strategy of explanation building, the researcher tries to
construct an explanation about the case by making statements about the cause or
causes of the phenomenon under study. This method can take several forms.
Typically, however, an investigator drafts an initial theoretical statement
about some process or outcome, compares the findings of an initial case study
against the statement, revises the statement, analyzes a second comparable case,
and repeats this process as many times as necessary. Note that this technique is
similar to the general approach of analytical induction.
For example, to
explain why some media websites are failing to generate a profit, a researcher might
suggest lack of managerial expertise as an initial proposition. But an
investigator who examined the situation might find that lack of management
expertise is only part of the problem, that inadequate market research is also
a factor. Armed with the revised version of the explanatory statement, the
researcher next examines the direct broadcast satellite industry to see whether
this explanation needs to be further refined, and so on, until a full and satisfactory
answer is achieved.
In time-series analysis, the
investigator tries to compare a series of data points to some theoretic trend
that was predicted before the research or to some alternative trend. If, for
instance, several cities have experienced newspaper strikes, a case study investigator
might generate predictions about the changes in information-seeking behaviours
of residents in these communities and conduct a case study to see whether these
predictions are supported.
- Report Writing The case study report can take several forms. The report can follow the traditional research study format— problem, methods, findings, and discussion—or it can use a non-traditional technique. Some case studies are best suited to a chronological arrangement, whereas comparative case studies can be reported from the comparative perspective. No matter what form is chosen, the researcher must consider the intended audience of the report. A case study report for policy makers is written in a style different from one to be published in a scholarly journal.
