连续反馈影响主导感的心理与脑电机制

his study aimed to investigate the impact of multiple sequential feedback on sense of agency (SoA) and its underlying cognitive mechanisms, bridging a gap in previous research that primarily focused on the influence of outcome feedback valences on SoA. Employing a simple gambling task (door task), the study manipulated the number of successive feedbacks and utilized explicit measure of SoA.A within-subject design of 2 (outcome valence: gain, loss) 8 (sequential feedback times: 1~8) was implemented. The study included forty college students (38 in the final analysis, with 20 females, aged 20.791.70), who voluntarily participated. During the experiment, participants guessed which of two doors contained a reward and selected the corresponding door by pressing a button. Feedback resulted in a gain of 10 for a correct guess and a loss of 10 for an incorrect one, with outcome feedback predetermined before the experiment, leading to one to eight different gain or loss sequences. Subsequently, participants rated their SoA after receiving feedback. Event-related potential responses to outcome feedback, focusing on stimulus-preceding negativity (SPN) in the outcome anticipation phase, and feedback-related negativity (FRN) and feedback-P3 (fb-P3) in the outcome processing phase, were recorded and analyzed.Behavioral results revealed a significantly lower SoA under loss conditions compared to gain conditions, aligning with previous findings. Sequential feedback demonstrated a cumulative effect on SoA, with continuous changes observed based on feedback results. Under gain conditions, the cumulative effect followed a quadratic polynomial model, with SoA increasing to a "change" point (the seventh gain feedback) and then stabilizing. Conversely, under loss conditions, SoA gradually diminished with continuous loss feedback, following a unitary linear model. Combining ERP and behavioral results, it was found that SoA under continuous loss was related to the amplitude of right frontal lobe SPN before outcome feedback. FRN amplitude was sensitive to the continuity of loss, being more negative under continuous loss compared to discontinuous loss and associated with a higher repetition rate of subsequent behavior selection. Moreover, continuous losses induced a significantly greater Fb-P3 amplitude compared to continuous gains.In conclusion, this study provides empirical evidence supporting the dynamic account of SoA, revealing differentiated developmental mechanisms in both behavior and ERPs under sequential gain and loss conditions. These findings shed light on how sequential feedback influences SoA and the characteristics of processing time.