Page 188 -

7.8 Data collection 175

data sources, you might want to rethink your design, adding additional data sources
or questions.

7.8.2 COLLECTING DATA

Once you have identified your data sources, you need to develop protocols for how
you will use each of them to collect data. For interviews, this will include the type of
interview, questions, and an interview guide (see Chapter 8). Similar approaches can
be used for examination of artifacts. Observations require you to specify the structure
of the tasks that will be performed and the questions that will be asked. Each data
source, in effect, becomes a mini-experiment within the larger case study, all tied to
the common goals of the study as a whole.
You should also develop a protocol for the case study as a whole. In addition
to the specific data sources and the procedures that you will use in examining each
of these sources, the protocol includes important details that are needed to conduct
the case study from start to finish. The case study protocol should start with an in-
troduction, including the questions and hypotheses. It should continue with details
of data collection procedures, including criteria for choosing cases, contact infor-
mation for relevant individuals; and logistical plans for each case, including time
requirements, materials, and other necessary preparations. Specific questions and
methods for each of the data sources should be included in the protocol. Finally, the
protocol should include an outline of the report that will be one of the products of
the case study (Yin, 2014).
Although this may seem like an excessive amount of overhead, effort spent on
careful development of a protocol is rarely wasted. The process of developing a clear
and explicit explanation of your research plan will help clarify your thinking, lead-
ing to a better understanding of possible shortcomings and challenges that may arise
during the study. Any problems that you identify can stimulate reconsideration and
redesign, leading to a stronger research plan.
A draft outline of your report serves a similar purpose. Constructing a report
before you collect any data may seem strange, but it's actually quite constructive.
Many of the sections of your report are easy to enumerate: your report will always
contain an introduction to the problem, a description of your questions and hypoth-
eses; an explanation of your design and how it addresses those questions; informative
presentations of data and analysis; and discussions of results. Within each of these
components there is substantial room for adaptation to meet the needs of each proj-
ect. An outline that is as specific as possible—even down to the level of describing
charts, tables, and figures to be used for presentation of data and analysis—will help
guide your design of the questions and methods that you will use to generate the
necessary data.
A case study protocol can be a powerful tool for establishing reliability (Yin,
2014). If your protocol is sufficiently detailed, you should be able to use it to conduct
directly comparable investigations of multiple cases—the protocol guarantees that
differences in procedures are not the cause of differences in your observations or