ABSTRACT
Interviews and group discussions produce many subjective truth claims: claims that people believe, feel, intend, value, and experience various things within a realm closed off to others. Subjectivity is the realm of privileged access. We depend on honest and accurate self-reports to learn about the subjective states of others. There are procedures that will strengthen stage three claims:
Use consistency checks on recorded interviews. A tape-recorded interview will generally be objectively sound. The objective features of these data concern only what was said, and a tape recorder usually captures that perfectly well. To analyze an interview transcript adequately, however, one must have some idea of whether the subject was being honest and accurate. One way of checking for honesty and accuracy is to determine whether or not the subject is consistent. If she says early on in the interview, “I hated my brother,” and then says much later on, “I admire my brother more than any other man I've ever met,” you may have problems in taking both statements at face value. The first part of conducting a consistency check is to look for any apparent discrepancies like this. If found, the next step is to go back to the subject and explain that you are confused about a couple of things she said. Show her the passages you have difficulty putting together and see how she explains them. She may acknowledge not being fully honest or clear at one point, or she may explain to you how the several sections in apparent disharmony are in fact consistent. One can both hate and admire someone, for example.
If the research design allows, interview the same subjects repeatedly. Repeatedly interviewing the same subject is likely to produce richer and more self-disclosing information than that produced in a single interview. You also build up a larger data set for conducting consistency checks.
Conduct consistency checks between observed activity and what is said in interviews. Frequently, your interviews will encourage a subject to talk about events you have observed or about matters related to what you have observed. You can then check the consistency between what the subject says and what you have in your primary record. When the subject talks about objective events, inconsistencies will reveal either a misperception on her part (assuming you used multiple recording devices and have a sound record of the events), a memory problem, or a lack of veracity You can go back to the subject if you suspect a lack of veracity and see what she has to say about the two conflicting accounts. Misperceptions, on the other hand, will tell you something about the interpretative schemes employed by the subject.
Use nonleading interview techniques. Leading can be very subtle. Simply asking a question that is too abstract, like “What are your views on standardized tests?” can be leading. I used a good deal of space in chapter ten on how to interview in nonleading ways.
Use peer debriefers for checks on possible leading. Peer debriefers ought to be invited to look through some of your interview transcripts early on, to pay close attention to what you said and asked. A good debriefer will be able to spot instances of leading on your part.
Use member checks. Member checking ought to be part of the interview process. Toward the end of an interview, show your subject some of the reconstructions you have conducted on field notes and invite commentary. You should also summarize the results of the interview, as you understand it, toward its end and invite challenges from the subject. With repeated interviewing of the same subject, it is possible to facilitate discussion of your analyses of previous interview material.
Encourage subjects to use and explain the terms they employ in naturalistic contexts. Subjects often will try to please the interviewer by adopting terms they think the interviewer favors. Discourage them from doing so. Ask them questions with terms you have learned from them and paraphrase what they say from time to time with their own terminology.
