Using accelerometer and gyroscope data in common mobile devices to assess credibility

An observational study examined the efficacy of using accelerometer and gyroscope (i.e., orientation) data from standard mobile devices to assess response credibility. Study participants played twenty rounds of a card game where they self-reported whether they won or lost each round. Participants received a base payment but could gain increased “bonus” payments by cheating. After the card game, participants were asked a “Yes-No” question whether they cheated while their orientation data was captured. We predicted that people concealing information about an adverse behavior (i.e., cheating), would result in differences in both absolute magnitude changes in acceleration and in rotation, as compared to truthful people. We developed six measures to capture absolute magnitude changes using orientation data. Supporting our hypotheses, cheaters had greater values of acceleration/rotation rates than non-cheaters. The implications of these results for future research and practice, as well as the limitations of this study, are discussed.