Assessing survey mode effects in the 2019 EP elections: A comparison of online and face-to-face-survey data from six European countries

Abstract

The quality of online survey data is sometimes seen as inferior compared to probabilistic face-to-face surveys. Extending existing research beyond the commonly studied cases, we analyse six European countries, comparing data from a representative online survey fielded for the 2019 EP elections to the European Social Survey 9 and EB 91.5. We assess the effects of survey mode in terms of representativeness compared to validated benchmarks, as well as with inferences about party identification and vote choice. While we do find some differences in the marginal distributions compared to sociodemographic and electoral benchmarks, the inferential models when using online survey data as compared to face-to-face surveys are not substantially different.