Page 186 -

7.7 Choosing cases 173

users. The additional credibility that comes from having appropriate participants is
referred to as external validity (Yin, 2014).
Multiple-case studies reduce concerns about external validity somewhat, as con-
sistent findings across your cases can be used to counter the argument that you are
describing some idiosyncrasy of your specific participants. However, these problems
reappear if you are attempting theoretical replication—members of each group must
both represent that group appropriately while differing from other groups in the ap-
propriate dimensions.
Sara's case study provides an instructive example of case selection. When reading
the paper, all we are told about Sara is that she is a blind college student. We are not
given any other details about her age, background, or socioeconomic status. However,
we can infer from the list of tasks—which includes activities such as organizing CDs,
cooking, and receiving text messages by cell phone—that she is fairly active and self-
reliant. In other words, as far as we know, she may be an appropriate participant for
a study of the workaround strategies used by people who are blind. We might not be
able to make generalizations that apply her results to other people, but that would be
true of any single participant. Furthermore, as the study was described as descriptive
and explanatory (Yin, 2014), the authors do not make any claims of generality.
Some case studies specifically seek out unusual, distinctive, or “edge” cases.
When studying antispyware or antivirus tools, you might argue that computer science
undergraduates are worth studying because you would look for an understanding of
how their domain expertise helped them approach challenges that would stop less
knowledgeable users. The Finnish study of virtual collaboration in a school setting
was conducted in a school that was chosen specifically because “the pedagogical
setting had several features that may be described as innovative” (Lakkala et al.,
2007). See the Extreme Cases sidebar for a description of a case study that specifi-
cally sought out an atypical set of participants in order to get a fresh perspective on
an established problem.
Some studies use critical cases—cases that are somehow particularly distinctive
or notable with respect to the problem that is being considered (Flyvbjerg, 2006).
For example, a case study examining the use of antivirus software by employees of
a large company might focus on a firm that required all staff members to complete
extensive training in the use of the tools in question. This required training makes the
firm a strong candidate for success: if antivirus software isn't used there, it might not
be used anywhere. Thus, the company becomes a critical case.
Still other strategies for identifying cases are possible. You might search for cases
that are most or least likely to exhibit behavior that you are interested in investigating
(Flyvbjerg, 2006).
If you find yourself trying to choose from a large pool of potential cases, con-
sider expanding your research agenda to include a screening survey (Yin, 2011).
A carefully constructed survey of potential participants can provide data that in-
forms your selection process. Such surveys might assess both the fit between the
participants and your criteria and the willingness of the participants to commit
their time and energy to the success of the study. Ideally, screening surveys stand