Large-scale social surveys are increasingly incorporating biomeasures to enhance social and health research. Yet collecting accurate anthropometric data outside clinical settings poses methodological and logistical challenges, especially in self-administered or mixed-mode designs. Mobile technologies offer promising alternatives, though evidence on their feasibility and performance in surveys remains limited. We address this gap by benchmarking the quality of biometric data gathered from a mobile app and assessing respondents’ willingness to share such information. Using data from the Understanding Society Innovation Panel, we examine three methods for collecting waist and hip measurements: (1) interviewer-assisted using tape measures, (2) self-administered using mailed tape measures, and (3) a smartphone app that generates a body outline from respondent-taken photos. We evaluate participation rates, non-participation bias, measurement consistency, and factors affecting measurement quality. Overall, participation was highest with interviewer-observed measurement (88.1%), followed by self-measurement (52.0%) and lowest with the app (18.4%). Nearly half (46.5%) of self-measurement non-participants cited non-receipt of the tape measure, while 36.1% of app non-users cited privacy concerns. Method selection varied systematically by age, education, and health status, influencing participation bias. In terms of measurement consistency, the interviewer-assisted approach yielded stable and reliable measurements, while self-measurement introduced discrepancies and under-reporting. The mobile app produced the least precise measurements, with a significantly more dispersed distribution, indicating greater variability and reduced accuracy compared to other methods. However, errors were largely driven by the conditions under which the photos were taken, limiting the app’s ability to detect body shape accurately. We conclude with practical recommendations for integrating biomeasures in large-scale surveys, highlighting trade-offs in cost, respondent burden, and accuracy—and the role of incentive design.

Co-authors

Professor Annette Jäckle; Dr Jonathan Burton, University of Essex; Professor Mick Couper, University of Michigan