Guidelines for the use of household interview duration analysis in CAPI survey management

This paper provides evidence based guidance for practical survey work, namely choosing interviewers and their workload. Analyzing a survey of 3568 households obtained through computer assisted personal interviews (CAPI) we find that interviewers learn considerably while the survey progresses. Time requirements for field work increase concavely with sample size which allows larger samples to be realized with a given budget than would be expected in planning such projects with simplistic cost estimates. We find a decrease of interview duration of almost 50 percent which translates into a significant increase of the average hourly wage the interviewers receive. These learning effects cease after around 20 interviews. Based on our results we recommend targeting interviewer training by age and technology-affinity of interviewers for CAPI surveys.