Linking interview speed and interviewer effects on target variables in face-to-face surveys

Research has explained interviewer effects in face-to-face surveys: in terms of question, respondent and interviewer characteristics. In this paper, we relate interview speed to interviewer effects. Three interview types are operationalized (slow, moderate and fast) based on the residual speeds from a model controlling for respondent characteristics. Data from the European Social Survey is used. Results show larger interviewer effects among slow and fast interviews than moderate interviews, although with variations between variables and countries. Assuming that moderate interviews are conducted under standardized interviewing, these results support the idea that deviations from standardized interviewing are linked to greater interviewer effects.