Case studies have typically been weakened by the conflict between two promises that are usually made to subjects being interviewed: (1) that the focus of the evaluation is the program, not the interviewee; and (2) that the interview is an opportunity for the subject to defend his or her practice to those who seek to judge it. These promises commit the evaluator to a conservative and insufficiently evaluative stance. In addition, traditional interviewing techniques do not serve the evaluative function. These improvements are suggested: (1) use a tape recorder for recording interview data, rather than a notebook; (2) concentrate upon the interviewee as a person rather than as a role incumbent; and (3) make it clear to the interviewee that the evaluative process is a shared task. (BW)