Striking the Balance: Understanding Cross-National Trends in Sample Composition within Self-Completion Web and Paper

The European Social Survey (ESS) is transitioning its mode of data collection from face-to-face interviewing to self-completion web and paper. Round 12 of the ESS, planned to begin in the autumn of 2025, will have samples randomly allocated to each mode, with Round 13 in 2027 being conducted entirely with self-completion. To date, the ESS has implemented over a dozen self-completion surveys across Europe that have seen a varying blend of web and paper questionnaire completions, with the web portions comprising of completions on different device types such as tablets, phones, and computers. Understanding the variation in sample composition between web and paper is vital to achieve a representative sample, and evidence from the ESS shows that some trends are not always applicable cross-nationally. For example, whereas paper completions have so far always brought in older, less educated respondents, in some countries they also tend to bring in a notable overrepresentation of women. The breakdown of web completions by device types also vary by country and are sensitive to particular methodological choices in designing invitation letters, such as including QR codes. The ESS has found that completions on mobile devices tend to bring in more younger respondents with notable country-specific differences. This paper will describe trends in sample composition by self-completion mode of completion (paper versus web) as well as by device type among web completions.