AN EVALUATION OF THE EFFECTS OF INTERVIEWER CHARACTERISTICS IN AN RDD TELEPHONE SURVEY OF DRUG USE.

Although a small number of studies are available that evaluate the effects of interviewer characteristics in substance use surveys conducted in person, none have done so using information collected via telephone interviews. We address this issue by examining the utility of social attribution and social desirability models for detecting the presence of interviewer effects in a large statewide telephone survey concerned with substance use. The specific outcome variables of interest were reports of lifetime and 18-month composite drug use. Analyses focus on the direct effects of individual interviewer characteristics (to assess social attribution) and a summary measure of interviewer-respondent similarity (to assess social distance) and employ random effects regression models to control for respondent clustering by interviewer. Results are most consistent with a social distance model and suggest that social distance between respondent and interviewer may decrease the probability of respondents reporting substance use behavior. [ABSTRACT FROM AUTHOR]