From question-behaviour effects in trials to the social psychology of research participation.

The 'question-behaviour effect' (QBE) has attracted much recent attention within health psychology, where it has also been referred to as the 'mere measurement' effect. There are other conceptualisations of similar phenomena in related disciplines. This paper explores the implications of the QBE for the safety of inferences about intervention effectiveness within the context of randomised controlled trials evaluating health behaviour change interventions. It draws attention to poorly understood mechanisms by which bias is introduced with conventional thinking about trial design and analysis. The threat to valid inference on intervention effectiveness posed by the QBE applies even when its effects are small and regardless of the specific content of the QBE. The nature of the resulting bias does not fit well within existing bias classification schemes, such as that proposed by the Cochrane Collaboration. The QBE is one possible consequence of research participation and it is suggested that the social psychology of research participation is very much underdeveloped. Possible future directions for health psychology research in this area are considered.