A Method of Measuring Interviewer Variability

Abstract

Survey researchers have long suspected that interviewer variability as well a sampling procedure may be a frequent source of error. In this article, the authors describe a method for measuring this variability; they suggest that it can be reduced by more intensive training and control of interviewers, by more careful question design, by the use of a larger number of interviewers, and by selecting interviewers more representative of the population being studied.