Influence of the Interviewer on the Accuracy of Survey Results

Abstract This paper reports results of a large scale study of the effect of interviewers on survey results. Where significant effects of the interviewer upon the results are found, the important factors appear to be (1) interviewer “resistance” to a given question—i.e., a tendency to omit or alter the question and/or to assume the answer; (2) relatively high ambiguity, “subjectivity,” or complexity in the concept or wording of the inquiry; (3) the degree to which additional questioning (“probing”) tends to alter initial respondent replies. The study also investigated the relationship of interviewer performance to interviewer characteristics as given by (1) scores on a number of tests; (2) by such personal characteristics as age, sex, occupation, education; and (3) by attitudes toward, and expectations of respondent reactions. At least one measure of poor quality of enumeration (number of improperly omitted Census entries) seems to show substantial correlation with age, with some test scores, and with expectations about respondent cooperation.