Thursday, November 10
Data Quality and Measurement Error
Thu, Nov 10, 1:30 PM - 2:55 PM
Hibiscus B
Exploring the Interaction Between Question Design and Interviewer Behavior on Measurement Error

The Effects of Interviewers and Personal Visits on Response Consistency (303514)

*Jonathan Michael Katz, Joint Program in Survey Methodology/U.S. Census Bureau 

Keywords: response consistency, interviewer variability, contact attempts

A consistent concern among survey methodologists has been the possibility that the interviewers may be influencing respondents’ answers in face-to-face interviews to survey questions. Nonresponse bias has become a more substantive issue with the decrease of response rate across a variety of modes. Nonresponse bias may be reduced by attempting to interview more difficult and hard-to-reach respondents. However, these tactics attempting to reduce nonresponse error may increase the risk of measurement error. This study attempts to evaluate how interviewers and number of contact attempts affect the “response consistency” in a series of social and political attitude items. Additionally there is an examination of whether the interviewer and contact attempts affect the measurement differently by type of attitude item. Mixed effects zero-inflated Poisson models and standard Poisson models are applied to analyze these two research questions. The data source retrieved for these analyses is Round 7 of the European Social Survey. Final results show that interviewers do contribute to the “response consistency” and that the variation is influenced by the type of attitude items. However, the relationship between contact attempts and “response consistency” is more enigmatic especially when comparing the associations between the types of attitude items.