An Investigation of Interviewer Effects on Measurement Error

• Activity Number: 616
• Type: Contributed
• Date/Time: Thursday, August 7, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Denize Barbosa*+ and Gabriele Durrant and Peter Smith
• Companies: University of Southampton and University of Southampton and University of Southampton
• Keywords: Multilevel models ; Interviewer effects ; Survey sampling

Abstract

Measurement error occurs when there is a discrepancy between an observed measure and its true value. Depending on the nature of the data collection, this discrepancy may be due to the imprecision of the instrument used. However, survey interviewers can also have an influence on measurement error since, in their presence, interviewees can be prone to give socially desirable answers that may differ from the truth. In this research we investigate interviewer effects on measurement error by applying multinomial multilevel models. We use data from the 2010 Norwegian sample of the European Social Survey (ESS) linked to administrative data. Our analysis is based on the joint distributions of variables of interest present in both sources of data. Preliminary findings suggest that the measurement error variable varies by interviewer.