Deception, risk-taking in social interactions: an ERP study

Abstract

Deception is a common behavior in normal life. However, little is known about the neural processing of making decisions to deceive. In the present study, we investigated the neural correlates associated with decisions on deception by using a modified Trust Game in which participants are asked to play as trustees and repay to investors, according to their own decisions, less than (i.e. deception) or equal to/more than (i.e. non-deception) required amount of repayment. Participants also have to consider the risk of being punished if their deceptions are detected. Our findings suggested two response-locked ERP components which reflected difference between deception and non-deception. One was before and during the stage of choice (-400 100 ms post response, Fig. 1A) with more positive amplitudes found most prominently at central-frontal sites for the decision to deceive than to respond non-deceptively. This difference was found most significant when the risk of being detected was high, and might reflect the processing of evaluation of the choice of deception. The other was found at central-parietal sites with the most significant difference appearing later (200300 ms, Fig. 1B) than the response. Deception was related with more positive amplitudes than non-deception, and this difference was not manipulated by the risk. The latter component might reflect the processing of numeric information of reward and the attention put on the reward kept for the participants. These results advance our understanding of the decisions made upon deception and suggest implications for application on lie-detection.