How can the Bayesian Truth Serum help people to live a healthier lifestyle?

People tend to under-report food intake and over-report physical activity. The misreporting of healthy behaviour is a problem because this leads to an inaccurate assessment of the current state of healthy behaviour in studies. This paper examines whether the Bayesian Truth Serum (BTS) can be used to elicit more honest and thoughtful self-reported responses to healthy behaviour questions. In order to estimate the effectiveness of the BTS in eliciting honest self-reports, an experimental design questionnaire was developed. Two hundred and twenty-four Dutch females completed the questionnaire, which included questions about general eating behaviour, fruit consumption, vegetable consumption, general physical activity, moderate physical activity, and intensive physical activity. A 2-by-2 between subject design was used for this study. Subjects were asked whether they were actively trying to lose weight or not, and in each group subjects were randomly assigned to the BTS group or the control group. It was found that the BTS did not elicit more honest and thoughtful self-reported responses. In contrast, results suggest that the BTS did elicit less honest responses to questions about physical activity. Furthermore, respondents who are watching their weight are more likely to report healthy behaviour. However, it was not found that the effectiveness of the BTS differed among weight watchers or non-weight watchers. Overall, the BTS did not provide more accurate self-reports in this particular study of healthy behaviour.