The Neural Basis Of Regret In Decision-making

Regret is an emotion associated with a decision that turns out badly. It is classically elicited by a comparison between the outcome of a choice (what is) and the better outcome of foregone rejected alternatives (what might have been). This comparison process is counterfactual thinking. Though regret has been deeply understood in behavior, its neural basis is still known little. It mainly embodies two aspects. On one hand, as far as regret's process concerned, there are many questions to be solved. The first one is what course has experienced from facing decision making outcome to producing regret. The second one is what features those processes are. One the other hand, as far as regret arousal concerned, there are many questions to be solved. The first one is whether regret arousal is affected by the magnitude of rewards and punishment and action effect. The second one is whether those influence will response on electrophysiological index.In order to explore above-mentioned questions, research conducts four series experiments by using wheel-of-fortune. Four experiments probe into regret electrophysiological characteristic through recording and analyzing event-related brain potentials(ERPs) induced by the decision-making outcomes. These experiments showed that: (1) after 200～300ms providing the feedback stimuli, decision-making error induces more negative wave than correctness, which labeled as feedback-related negativity(FRN); Decision-making correct induces more positive P300 than error after 300～400ms of the feedback stimuli onset; Decision-making correct induces more positive late positive component (LPC) than error after 500～900ms of the feedback stimuli onset, and this phenomena show significant right hemispheric specialization. (2) FRN is only sensitive to decision-making correct or error and insensitive to gain-and-loss or its magnitude; P300 is insensitive to the valence of feedback stimuli, but sensitive to gain-and-loss and its magnitude; LPC is insensitive to the magnitude of gain-and-loss, but the relationship amongst LPC,decision-making correct and gain-and-loss is very complicate, because this three interact with each other. The sensitivity of LPC on decision-making correct or error is seemly modulated by gain-and-loss, in the same way the sensitivity of LPC on gain-and-loss is seemly modulated by decision-making correct or error. (3) As far as small gamble concerned, error induces more negative FRN than correct, and correct induces more positive P300 than error, but LPC has no significant differences between correct and error. (4) When discussing the relationship between gamble's magnitude and FRN or P300 or LPC, researches find out that: Firstly, large gamble's amplitude of FRN is higher than small gamble, which only consist in Fz and FCz; Secondly, large gamble's amplitude of P300 is higher than small gamble, which only consist in Fz and FCz; Thirdly, the amplitude of LPC has no significant differences between large and small gamble. (5) Reject/error induces more positive LPC than accept/error, which is consistent with action effect There is no significant differences on the amplitude of LPC between accept/correct and reject/correct.The results above show that: (1) Under the paradigm of wheel-of-fortune, FRN mainly process the significant information provided by feedback stimuli, which indicate subject's decision is correct or error. Moreover its amplitude is partly sensitive to grade of reward-and-punishment. (2) P300 mostly encodes magnitude of reward-and-punishment. Different reward-and-punishment affect the intensity of motivation and emotion. (3) LPC mainly reacts on emotional experience which is induced by correct and error. Due to those three show inconsistent on such aspects, researcher speculate that they have different functions during procession feedback stimuli. The functional dissociations is related to regret processing time, and indicates the time course facing decision making outcome to producing regret.