Date of this Version
Presented at “Interviewers and Their Effects from a Total Survey Error Perspective Workshop,” University of Nebraska-Lincoln, February 26-28, 2019.
Interviewers systematically speed up over the field period of a survey as they conduct interviews (Olson and Peytchev 2007; Olson and Bilgen 2011; Kirchner and Olson 2017). Competing hypotheses for this increase in speed is that interviewers learn from previous interviews, changing their behaviors accordingly, or that they change behaviors in response to who the respondent is, including both respondent’s fixed characteristics and their response propensity. Previous work (e.g., Kirchner and Olson 2017) has failed to completely explain this learning effect, even after accounting for a wide range of measures of each of these hypotheses. However, prior work has not examined how actual behaviors during an interview are related to interview length and whether different interview lengths can be explained by different types of interviewer behaviors. This paper uses data from two telephone surveys with extensive information on interviewer behaviors to attempt to explain the within-survey interviewer experience effect. Results indicate that interviewer behaviors do change over the field period (e.g., reductions in inefficient interviewer behaviors) and that interviewer behaviors are related to interviewer length, but that interviewer behaviors do not fully explain decreases in interview length over the field period. Implications for future research and for practice are provided.