As response rates to face-to-face surveys continue to fall while costs increase, surveys are increasingly switching, in whole or in part, to online data collection including an increasing use of mobile ‘smart’ digital devices for completing questionnaires. This existing trend is likely to be accelerated by the coronavirus pandemic, which has led to the suspension of face-to-face interviewing around the world. Many surveys that have been forced online by the pandemic are unlikely to return to face-to-face interviewing, even when this becomes possible. There is therefore a pressing need to improve our understanding of how data collection in online environments can be delivered to optimize data quality.
Online surveys have many advantages over traditional modes of interview such as, notably, cost savings, increased frequency of interviewing in longitudinal studies and the potential to reach groups that are normally harder to reach. The earliest adopters of online surveying used non-probability sampling approaches and the vast majority of online surveying continues to use these kinds of sample designs. Increasingly, however, researchers are seeking to implement web surveys using random probability designs with more and more online probability panels appearing in countries around the world. Implementing probability sampling for online surveys raises a number of difficult methodological challenges such as within household respondent selection, identifying ‘fake’ respondents, and achieving satisfactory response rates, to name but a few. This will be the focus of the research funding by this studentship, and the successful candidate will undertake a programme of empirical research focusing on these kinds of challenges.
The research will focus primarily, although not necessarily exclusively, on the Kantar Public Voice probability panel. Kantar Public Voice is one of only two random probability panels in the UK. It currently comprises over 10,000 panel members who were recruited through a combination of face-to-face and postal (push-to-web) invitations. Panel members are invited to complete surveys on a monthly basis and are given a monetary incentive for each survey completion. The studentship will have access to the wealth of paradata emanating from the data collection process, there will also be opportunities for implementing experimental designs as part of future rounds of data collection. The full programme of work to be undertaken for the PhD will be developed collaboratively during the first year of the studentship between the student and the supervisors, which will include a senior methodologist from Kantar Public.