Multivariate assessment of interviewer errors in a cross-national economic survey and the role of fieldwork institutes

Abstract Interviewers have long been identified as a source of error in face-to-face surveys. However, previous studies have typically focused on a single source of interviewer error and single-country cross-sectional surveys. We extend this literature by investigating interviewer errors from multiple dimensions in the Oesterreichische Nationalbank Euro Survey, a cross-national survey conducted annually in 10 Central, Eastern, and Southeastern European countries. Using data from 10 rounds (i.e. 100 country-years), we apply several data quality indicators on various dimensions of interviewer error and investigate country-years with particularly exceptional patterns. To combine the indicators, we use a multivariate tree-based outlier detection method (isolation forest) that flags country-years and interviewers with outlying values and combine it with methods from the interpretable machine learning literature to identify the respective exceptional feature values. Lastly, we document the effects of interviewer errors on the bias and variance of survey estimates. In several instances, our results identify fieldwork institutes and supervisors rather than interviewers as the main source of error.