Interviewer effect on responses to a questionnaire relating to mood.

A community mental health assessment questionnaire relating largely to depressed mood was administered during 1972 to 1,212 respondents in Washington County, MD, by six interviewers. Analysis of 15 psycho-social tests showed that responses obtained by one interviewer differed significantly from responses obtained the other five. Suggestions for minimizing interviewer effects include 1) selection of interviewers with similar characteristics and backgrounds; 2) adequate training and periodic field assessment of interviewer performance; 3) simplification of questions and reduction in the number of possible responses per question; and 4) allocation of various types of subjects to interviewers as uniformly as possible.