Functional Heterogeneity Of Perceived Control In Feedback Processing

Aim:Perceived control is a fundamental psychological function associated strongly with human well-being,and it can either boost positive affective or buffer negative affective.Recent studies have demonstrated that both the reward positivity(Rew P)and the feedback P300(fb-P3),which are relevant electrophysiological indexes during feedback processing,are larger when outcomes are delivered after voluntary choices than passive choices.However,none of these studies has addressed the choice effect in processing feedback with different valences.Here,using choice as a vehicle for perceived control,we focused on the electrophysiological mechanisms underlying perceived control in processing of positive and negative feedback.Methods:Thirty-six normal participants were recruited in the study to perform a simple choice task while their electroencephalography(EEG)data were recorded.The participants were asked to make the correct decision between the two doors in order to get the points,and they could receive positive(gain)or negative(nongain)feedback following choices made either by themselves(high control condition)or by a computer(low control condition).Following the choice task,participants'choice desirability was measured through their choice behavior on a choice-preference task.All statistical analyses were performed using IBM SPSS(v19.0).Firstly,it includes comparison of reaction time in different conditions within EEG tasks,comparison of participants'report scores between different cues,and statistical analysis of proportion of choice preference.Secondly,it also includes the influence of the choice availability and feedback valence on each electrophysiological signal,and the analysis of the relationship between the groups obtained according to the participants'proportion of choice preference and the electrophysiological indexes.Result:1.Participants took a longer time to select a door in the choice condition(744.90±188.41 ms)than in the no-choice condition(658.37±132.42 ms),t(1,35)=4.46,p<.001.Moreover,participants perceived a higher level of control for choice trials(7.17±1.78)than for uncertainty trials(5.97±1.72),which was in turn higher than no-choice trials(4.47±2.52),F(2,70)=21.86,p<.001,?_p2=.38.2.Data of ERP components(Rew P and fb-P3)and EEG power(delta and theta)were analyzed separately with a Choice Availability(choice,no-choice)×Feedback Valence(gain,nongain)ANOVA.For the Rew P,there was a significant interaction between choice and valence,F(1,35)=11.88,p=.001,?_p2=.25.Post hoc comparisons revealed that although the choice effect was significant for both outcomes,it was more pronounced for gain outcomes(?M=3.30?V,p<.001)than for nongain outcomes(?M=1.40?V,p<.01).In contrast to the Rew P,for the fb-P3,the interaction between choice and valence was not significant,F(1,35)=2.07,p=.16,?_p2=.06,indicating that they were encoded independently during the time window of the fb-P3.Moreover,the interaction between choice and valence was significant,F(1,35)=5.16,p=.029,?_p2=.13.Post hoc comparisons revealed that although theta power exhibited a choice effect for both gain and nongain outcomes,it was more pronounced for nongain outcomes(?M=1.18 d B,p<.001)than for gain outcomes(?M=0.62 d B,p<.01).In contrast to theta power,for the delta,the interaction between choice and valence failed to reach significance,F(1,35)=1.65,p=.207,?_p2=.05,indicating independent effects of choice and valence on delta power.3.The mean percentage of choice preference was 61%.Based on this,the participants were divided into a self-preference group and a computer preference group,and then repeated-measure ANOVA was performed.We found that the interaction between choice and group was significant for fb-P3,F(1,34)=4.68,p=.038,?_p2=.12.Moreover,delta power exhibited a significant three-way interaction among choice,valence,and group,F(1,34)=7.70,p=.009,?_p2=.19.Post hoc comparisons revealed that the self-preference relative to computer-preference group exhibited enhanced delta power for choice-gain outcomes(2.85 vs.1.75 d B,p=.047)but not for other types of feedback(p>.120).4.Following correlation analyses revealed that the proportion of choice preference was positively correlated with fb-P3 amplitude in the choice condition,r(36)=.37,p=.027,and delta power in response to choice-gain outcomes,r(36)=.34,p=.046.Conclusions:1.The Rew P findings reflect that the perception of control can be rewarding in and of itself,whereas theta power results suggest that the feeling of control can increase the need for control following receipt of negative feedback.2.Furthermore,both the fb-P3 and delta power findings indicate that perceived control enhances the salience of outcome irrespective of valence.3.These results suggest functional heterogeneity of perceived control in feedback processing as diverse as magnifying the reward signal,enhancing the need for control,and increasing the salience of outcome irrespective of valence.