The Smartphones Study: An Analysis Of Disciplinary Differences In Research Ethics Committee Responses To Phone App-based Automated Data Collection

Abstract

Digital devices provide a means to collect vast quantities of data relevant to health and social research with minimal respondent burden, but the automated harvesting of data without active participation raises ethical issues. We aim to present findings from research on ethics committees' opinions of a project that used a smartphone app to collect location, calls and text, leisure activities, illicit drug use and photographic data from schoolchildren. A research protocol for this school based study of substance use and social networks using a phone app for data collection was sent to 13 ethics committees and the research governance office in Queen's University Belfast. The committees were asked to give the protocol scrutiny as if it were being submitted as a research study for approval. The protocol, and the proposal to submit it to other committees, was approved by the Sociology ethics committee. Thematic content analysis was conducted based on the returned comments. Out of 13 ethics committees invited, five provided responses; five were humanities or physical sciences where the project was of no relevance, and Medical, Nursing, and Geography committees did not respond. All responding committees gave a favourable opinion but requested further information or minor changes. Emergent themes included: providing participant information regarding automated data collection; anon-ymising geographical and call information; awareness of digitised & online identity; and data security. There were large variations in the scale and content of responses, with the education research ethics committee providing the most varied and largest number of issues. Phone app based data collection, even with a high level of invasiveness and in relation to sensitive topics is generally viewed favourably. The variation in ethics committee responses suggests that applied ethics are greatly influenced by reviewers' expertise and training rather than by adherence to overarching ethical principles.