Is There More Than the Answer to the Question? Device Use and Completion Time as Indicators for Selectivity Bias and Response Convenience in Online Surveys

Abstract

The main objective of the German Emigration and Remigration Panel Study (GERPS) is to establish a longitudinal data set that provides information on life trajectories of international migrants. However, a large amount of paradata were also collected in order to obtain meta-information on respondents’ survey participation. This auxiliary information can help to optimize data quality at all stages of the survey process. By continuing the existing discussion in the field of online surveys, this chapter pursues a twofold objective: it reflects device usage (mobile vs. computer) and elucidates determinants of device choice. In particular, it analyses whether selectivity effects due to respondent’s device choices bias the sample. Moreover, this chapter investigates differences in response time between devices to detect differences in response burden. The analysis of response burden differences by device is an important issue, since an increased device-specific response burden can be a predictor of actual and further panel dropouts. In both device-specific selectivity and survey burden, only slight differences were found between mobile and desktop devices. Using these data, the following paper addresses the need to analyse potential sources of survey error and provides evidence that GERPS data do not appear to contain noteworthy bias attributed to device usage.