Performing Subjectivist Studies in the Qualitative Traditions Responsive to Users - Evaluation Methods in Biomedical Informatics - page 282

notes and eye contact can be retained. Human subjects committees often
have special requirements for interviews that are to be recorded. For
example, these committees will usually require that the consent form
specify where the actual tapes will be stored and for how long. As part of
further preparation, the interviewer should gather information about the
interviewee ahead of time and develop an interview guide listing several
main questions and many subquestions that can probe for answers if the
interviewee does not spontaneously respond. Some questions may be asked
of everyone who is interviewed, while other questions will be reserved for
specific types of individuals. It is absolutely critical that the interviewer be
a good listener and avoid making any value judgments during the interview.
Judicious use of silence can often stimulate responses. Interrupting is to be
avoided. Many investigators prefer to allocate a full hour to conduct an
interview of this type; however, shorter interviews can be useful if they have
a well-defined focus or if the informant is being contacted on multiple occa-
sions. After the interview, the tapes should be transcribed by someone with
experience in capturing nuances such as laughter and sighs; it usually takes
3 hours or longer to transcribe 1 hour of tape. Further details about con-
ducting semi-structured interviews are provided in Appendix B.
Group interviews, or focus groups, differ fundamentally from the one-on-
one interviews discussed above. They should not be viewed as a way to
gather more interview information in less time. Focus groups are not easy
to do well and, if several participants speak at the same time, audiotapes of
focus groups can be hard to transcribe. When run properly, however, focus
groups have distinct advantages. In particular, useful synergy among the
participants can develop. When this happens, participants build on the
thoughts of one another to generate new insights or more accurate recol-
lections of past events. (It is interesting to listen to members of a focus
group correct one another's personal recollections of a past event, until a
more accurate consensus develops.) Focus groups can employ many of the
same types of questions as those used in one-to-one interviews. Gathering
a group of up to 10 informants over pizza at lunchtime can generate lively
narratives. The moderator needs an interview guide and must set some
fairly strict ground rules to discourage participants from interrupting or
monopolizing the discussion. An assistant moderator can take general notes
that specify who is speaking (which will help the transcriptionist), and can
also manage the audiotaping. 18
Consensus or expert panels can also be treated like group interviews. The
advantage here is that informants may be together for extended periods of
time, perhaps for days, so rapport and synergy can build over time and
settled opionions of the full group can be generated. Audiotaped transcripts
of these discussions can be formally analyzed as qualitative data sources.
Often the conversation leading up to an agreement is filled with vivid
stories and examples, and worthy of capture and analysis. 19