Giorgio Piccitto , Bocconi University
Aart C. Liefbroer, Netherlands Interdisciplinary Demographic Institute (NIDI)
The decreasing response rates and increasing costs of face-to-face surveys and the increasing internet penetration make it attractive to use web surveying as an addition to or replacement for face-to-face surveys. The Generations and Gender Survey, one of the most widely used resources of comparative research among demographers plans to move to web surveying as its major method in future waves of data collection. This paper uses data from a pilot study conducted in Germany and Croatia to examine the consequences of moving to web for estimates of central variables of interest to demographers and family scholars and for estimates of the strength of associations within theoretical models. Given that social desirability bias is assumed to be lower in web than in face-to-face mode, we expect less social desirable answers to questions that are sensitive to one’s self-representation in web than in survey mode. We do not expect large differences between both modes in the strength of associations between variables in multivariate models. These expectations are tested on a range of variables and models of interest to demographers and family scholars.
Presented in Session P3. Poster Session Migration, Economics, Environment, Methods, History and Policy