Interviewer Effect on the Survey Data in Lithuania: Analysis of Three Surveys

  • Ainė Ramonaitė
Keywords: survey quality, interviewer effect, measurement error, intra-class correlation coefficient, political attitudes

Abstract

The article analyses the interviewer effect on the data of three public opinion surveys on political attitudes of the population conducted in Lithuania. The study discusses why in international comparative studies Lithuania stands out for its extremely high interviewer effect, which raises serious doubts about the reliability and suitability of the data for analysis. The article, first, reviews the reasons for the interviewer effect and the methods of its measurement and, second, presents the results of multilevel modelling. The analysis of surveys conducted by three different public opinion research agencies reveals that the interviewer effect varies significantly depending on the research agency. The hypotheses on the differences in the interviewer effect related to the nature of the questions were not confirmed, but it was found that the interviewer effect was greater on more abstract and complex questions. In the conclusions, the recommendations for researchers working with surveys on how to control the interviewer effect are provided.
Published
2020-12-01
Section
Articles