Brain Dynamics Of Generalized Trust Game And Its Individual Differences

Human societies are probably unique in the extent to which generalized trust is present to some degree in all interpersonal interactions. Generalized trust refers to one’s willingness to place one’s resources at the disposal of another party, sometimes without adequate information for judging the other person’s motives, intentions, and actions. Widely regarded as a paradoxical phenomenon, generalized trust involves the juxtaposition of human’s loftiest ambition with their deepest fears. On the one hand, as a fundamental social signaling mechanism, generalized trust plays an important role in the development and maintenance of well-functioning relationships. On the other hand, generalized trust presupposes risk and uncertainty when sufficient information is missing to judge other’s trustworthy. The decision of whether and when people trust one another becomes a social dilemma which impacts our way of life, and has been investigated by psychologists, economists, sociologists, neurophysiologists, as well as scholars in other academic fields for decades.Using the Trust Game, researchers have explored cognitive neural mechanisms of interpersonal trust and have proposed four theories including the betrayal aversion theory, social norm theory, moral norm theory, and the default behavior model. Neuroimaging studies have indicated certain brain regions to be differentially associated with the decision to trust, including medial prefrontal cortex, caudate nucleus, amygdala, and insula. According to the information processing theory, human’s decision-making can be divided into three stages including choice evaluation, response selection, and feedback processing. Decision-making is dynamic rather than static, and the psychological activities at different time-course would leads to totally different results. Moreover, most of the neuroimaging researches in this field have mostly focused on the functional location of different brain networks. However, it is still unclear about the temporal and spectral features of the investor’s decision-making and outcome evaluation during Trust Game.In order to address these questions, this dissertation paired the Trust Game tasks with event-related potentials (ERPs) and event-related oscillations (EROs) to investigate the time course and neural oscillation courses of generalized trust, and how risk attributes and individual differences affected the processing of decision making and outcome evaluation during Trust Game. This dissertation was composed of four electroencephalographic (EEG) studies. Study 1 explored the time course and neural oscillation courses of trust decision during a single-shoted Trust Game. To investigate the time course and neural oscillation courses of decision making under risk in social and non-social contexts, Study 2 recorded the EEG data from participants while they played the role of trustor and should make a decision to trust or not trust the human counterpart or a computer partner. Study 3 using the ERPs/EROs technique investigated how social risk modulated the neural activities of decision making phase and outcome evaluation phase during a single-shoted Trust Game. Study 4 explored how social value orientation modulated the neural activities of decision making phase and outcome evaluation phase during Trust Game. The results were as follows:1. Behavioral results revealed that participants made trusting decisions more often than chance. Electrophysiological results found that distrusting choice induced a more negative N2 and a less positive P3 than did trusting choice. Spectral analysis results showed the beta-band (18～28 Hz,250～350 ms) power for distrusting choice was significantly larger than trusting choice, which reflected inhibitory control. Moreover, trusting choice after a negative outcome elicited a significant higher beta-band (15～20 Hz,150～300 ms) power than did trusting choice after a positive outcome, while the beta-band power for distrusting choice after a negative outcome was similar with distrusting choice after a positive outcome.2. Behavioral data suggested that participants selected more trusting decision when interacting with human partner than with computer partner. In the decision making phase, distrusting choice induced a more negative N2 and a larger beta-band (13-15 Hz,250-350 ms) power thandid trusting choice when interacting with human partner whereas it did not show differential responses to different choices when interacting with computer partner. During the outcome evaluation phase, the differentiated feedback related negativity (dFRN) was increased when interacting with human partner than with computer partner. Moreover, gain feedback induced a more positive P300 and a larger beta-band (16～24 Hz,200～300 ms) power than did loss feedback when interacting with human partner whereas it did not show differential responses to different feedbacks when interacting with computer partner.3. Behavioral data indicated that the average trust rates for low social riskers were lower than for high social riskers. In the decision making phase, distrusting choice induced a less positive P2, a more negative N2, a less positive P3 and a smaller beta-band (14～16 Hz,350～450 ms) power than did trusting choice for high social riskers whereas it did not show differential responses to different choices for low social riskers. During the outcome evaluation phase, the FRN, when calculated as the difference between loss feedback and gain feedback, was enhanced for low social riskers than for high social riskers. Moreover, gain feedback induced a larger beta-band (16～22 Hz,250～350 ms) power than did loss feedback for high social riskers whereas it did not show differential responses to different feedbacks for low social riskers.4. Behavioral data indicated that the average trust rates for proselfs was lower than for prosocials. In the decision making phase, distrusting choice induced a more negative N2 than did trusting choice for proselfs whereas it did not show differential responses to two options for prosocials; the beta-band (26～30 Hz,220～300 ms) power was larger in response to trusting option compared to distrusting option for prosocials whereas the main effect of decision option was insignificant for proselfs. During the outcome evaluation phase, a larger FRN and a larger theta-band (5-8 Hz, 350～450 ms) were increased in response to loss feedback versus gain feedback for proselfs than for prosocials.In summary, these behavioral results highlighted the ingrained norm of cooperation and trust in modern society and indicate that the human generalized trust behavior were not only influenced by risk attribute, but also modulated by social risk and social value orientation. The psychophysiological data suggest that,1) in the decision making stage, when making a decision to trust or not to trust the counterpart, distrusting choice induced the larger N2 and an increase of beta power over frontal areas than trusting choice, which maybe reflect frontal inhibitory processes and cognitive control; 2) during the outcome evaluation stage, people firstly evaluated the violation and discrepancy of the cooperation norm which reflected by the FRN and theta oscillatory activity, then evaluated the outcome’s motivational/affective salience and allocated attention resources to different outcomes which reflected by the P300 and beta oscillatory activity. The spatio-temporal and spectral feature of decision making process and outcome evaluation process during interpersonal trust situation is not only influenced by task situation, but also vary as a function of individual differences (e.g. social risk and social value orientation). The studies of this dissertation firstly highlight the ingrained norm of cooperation and trust in modern society, and provide preliminary spatio-temporal and spectral evidence for human’s generalized trust behaviors.