|
Siyang Kong , Netherlands Interdisciplinary Demographic Institute
Arieke J. Rijken, Netherlands Interdisciplinary Demographic Institute
Xiao Xu, Netherlands Interdisciplinary Demographic Institute (NIDI)
Web surveys are becoming a popular way of collecting data for scientific purposes. A key feature of web surveys is that they can be completed using a range of devices. We have yet a limited understanding of how preferences for using certain device differ between people and how this selection process may interfere with survey quality, as indicated by interview duration and survey completion. In the second round of the Generations and Gender Survey (GGS-II), many countries have collected data using web primarily. By combining paradata generated during data collection and microdata, we examined the patterns of device use by respondents’ demographic and socioeconomic characteristics in six GGS-II countries. The results of logistic regressions suggest that female, younger, lower educated and economically disadvantaged people are more likely to use smartphones than PCs to answer the online survey. While looking into interview duration and survey completion, we found no substantial difference between phone users and PC users. These findings indicate that the multi-device characteristic of web survey is not as problematic as previous research has suggested. However, we also found that the patterns of device use vary greatly across countries, suggesting a potential issue in the effectiveness of cross-country comparison in large online surveys.
Presented in Session 28. Flash session New and Critical Perspectives on Data Collection and Measurement