Annette Jäckle, Carina Cornesse, Alexander Wenz, Mick P. Couper
Measuring expenditure with a mobile app: Do probability-based and nonprobability panels differ?

Previous studies have shown that, while nonprobability panels are much cheaper to maintain than probability-based panels, estimates are often less accurate, even when taking account of differences in socio-demographic panel sample composition. In this paper we examine expenditure data collected with a mobile app over a period of one month in a probability-based and a nonprobability panel. We find differences between the app study samples in who participates in the mobile app study and in the expenditure captured with the app, even after accounting for differences in panel sample composition, but no differences in how participants used the app.