عوامل مرتبط با فیزیولوژیکی روانی همکاری بین فردی و خشونت

Mimicking real world situations, the Chicken Game allows scientists to examine human decision-making when the outcome is not entirely within one person's control. In this social dilemma task, two players independently choose either to safely cooperate with, or riskily aggress against, the other player, and the unique combination of their choices specifies the outcome for each. Coupling the Chicken Game with psychophysiological measures, we confirmed our two hypotheses: that an individual perceives an outcome as most negative when she chooses to cooperate and the other player violates that trust and aggresses, and that motivational salience of an outcome is greater when an individual chooses to aggress and when she gains money. Collectively, the data demonstrate the utility of pairing true social dilemma tasks like the Chicken Game with psychophysiological measures to better understand decision-making.

The issue of reducing national debt plagues countries internationally and is governed by a unique webbing of risk-taking by, and payoffs for, independent parties. In the 2012 Greek elections, for instance, the political faction elected determined whether or not Greece would accept the austere bailout terms of European Union. Electing a left-wing party would have led Greece to renounce her debts, refuse the bailout plan, and leave the euro zone; instead, the election of a right-wing party has caused Greece to acquiesce with the bailout plan, maintaining her place in the euro zone. Independent of Greece's decisions, the European Union had the choice to alter the bailout plan in favor of Greece's financial welfare, but instead chose to maintain the strict terms to benefit its own financial state. As both Greece and the European Union were faced with choices, it was the combination of their independent decisions that collectively determined the financial outcomes for each. Like governments faced with economic crises, individuals regularly confront decisions involving various degrees of risk and reward. Whether deciding how to merge into oncoming traffic or if there is time to zip quickly across a one-lane bridge before an approaching car, the final outcomes of many decisions are not determined solely by your choice but instead by the unique combination of your choice and that of someone else. Do you select the safer, more cooperative option in which you risk being taken advantage of? Or, rather, do you choose to aggress and dare the other person to let you come out ahead, despite the risk that you both might end up with a bad outcome? Social psychologists and economics refer to this dilemma as the Chicken Game (Rapoport & Chammah, 1966). Social decision-making games such as the Chicken Game mimic real world situations by allowing us to examine human decision-making in situations where the outcome is not entirely within one's control. The Chicken Game presents players with a true social dilemma, as two players independently choose to either cooperate with, or aggress against, the other player. Choosing to aggress permits maximal personal gains if the other player cooperates; however, both players choosing to aggress earn each the worst outcome. For the player who cooperates when the other aggresses, her reward is slightly better than when both choose to aggress. In contrast, if both players cooperate, they each have moderate gains, which if consistently chosen by both players, would result in the best overall outcomes for each ( Fukui et al., 2006). Table 1 provides the relative payoffs of the different social interactions and hence depicts the tension that the game facilitates between cooperative and competitive impulses. Table 1. Payoff matrix in the Chicken Game. The outcome (in Yuan) for the participant (Player 1) is displayed outside parentheses, and the outcome of the opponent (Player 2) is displayed within parentheses. Player 2 Cooperate (C) Aggress (A) Player 1 Cooperate (C) 10 (10) −10 (30) Aggress (A) 30 (−10) −30 (−30) Table options Pairing social-dilemma tasks like the Chicken Game with psychophysiological measures permits insight into human decision-making. Although the current study examines event-related potential (ERP) correlates of cooperation and aggression, functional magnetic resonance imaging (fMRI) literature informs this work. Research on the neuroscience of cooperation and aggression has employed social decision-making tasks like the Trust Game (TG), Prisoner's Dilemma Game (PDG), and Ultimatum Game (UG). This literature suggests that decisions to cooperate or compete are guided by both bottom-up and top-down cognitive processes. For example, the caudate nucleus appears to function in learning reward values of stimuli (Montague, Dayan, & Sejnowski, 1996) and tracking an opponent's decision to reciprocate or not reciprocate cooperation in the TG and PDG (King-Casas et al., 2005, Rilling et al., 2002 and Rilling et al., 2004). In these ways, the caudate indexes social prediction errors and guides decisions about one's own reciprocity (Rilling, King-Casas, & Sanfey, 2008). Unreciprocated cooperation in particular is associated with activation of the insula, which alongside evidence that the insula is implicated in aversive conditioning, suggests that it may mark negative social interactions in an effort to learn to avoid them in the future (Rilling et al., 2008 and Seymour et al., 2004). However, cognitive control processes may also override the networks implicating these structures and exert a top-down influence on decision-making (Rilling et al., 2008). For example, the dorsolateral prefrontal cortex, involved in goal maintenance and executive functions (Miller and Cohen, 2001 and Wagner et al., 2001), appears to also be involved in the inhibition of the emotionally driven response of the insula, in an effort to rationally accept unfair offers and increase monetary gains (Sanfey, Rilling, Aronson, Nystrom, & Cohen, 2003). Similarly, individual differences in decisions to trust others captured by the amygdala, motivation to voluntarily donate money is represented in the ventral striatum, and norm-driven responses to punishment observed in the lateral orbitofrontal cortex have also been demonstrated to exert top-down influence (Rilling et al., 2008). The anterior cingulate cortex (ACC) is also critically involved in the processes of cooperation and competition. In addition to increased DLPFC and insula activity following unfair offers in the UG, individuals also had increased ACC activation (Sanfey et al., 2003). With a cognitively oriented dorsal region (dACC) and an affectively oriented rostral region (rACC), Sanfey and colleagues interpreted the ACC as representing the conflict between rational and emotional responses to unfair offers, such that increased ACC activation to unfair offers reflects augmented negative emotional responses. Additional research provides support of the ACC's sensitivity to the affective properties of social pain. For example, Eisenberger and Lieberman (2004) found increased activation of the ACC following social exclusion by peers, and Burklund, Eisenberger, and Lieberman (2007) found that individuals with heightened sensitivity to rejection had increased ACC activation in response to disapproving facial expressions. In the same way that offers in the UG elicit cognitive and emotional responses, the Chicken Game provides outcomes that research purports are evaluated similarly by the ACC. Numerous studies provide evidence that the ACC is associated with outcome evaluation, such that there is greater ACC activity for more negative outcomes (Behrens et al., 2007, Quilodran et al., 2008 and Walton et al., 2004). Outcome evaluation is, in fact, an important step in adaptive decision-making (Paulus, 2005), as encoding the outcomes of previous actions helps an individual to determine how future decision-making behavior should be altered (Platt, 2002). In support, it appears that the evaluation of outcomes in the ACC is critical in the process of selecting motor actions to guide future behavior (Quilodran et al., 2008 and Shima and Tanji, 1998). As cooperative and competitive tendencies are woven in with the complex social environments addressed in the Chicken Game, exploring how outcomes are evaluated provides the opportunity to understand how decisions are made. Psychophysiological research purports that the brain has developed special mechanisms to quickly assess the valence and magnitude of outcomes, as well as their subjective, motivational significance (Leng & Zhou, 2010). Two event-related potentials (ERP) components index such aspects of outcome evaluation: the feedback-related negativity (FRN) and the P300. The FRN is a negative-going component that is maximal over fronto-central recording sites between 200 and 300 ms after feedback presentation, and whose neural generator is evidenced to be the ACC (Gehring and Willoughby, 2002 and Holroyd and Coles, 2002). It is consistently larger for more unfavorable outcomes, which alongside evidence of the ACC's role in representing affective pain, suggests that the FRN reflects whether the individual achieved the desired feedback (Holroyd et al., 2006 and Sanfey et al., 2003). A second ERP component indexing evaluations of decision-making is the P300, a positive-going component that peaks at parietal electrodes between 200 and 600 ms after feedback presentation. The generation of the P300 appears more widely distributed, although evidence converges across methodologies to suggest that the ACC and parietal cortex may play critical roles in its generation (for review, see Linden, 2005). The P300 is most consistently thought to reflect the processes of attentional allocation and motivational salience (Gu et al., 2010, Hajcak et al., 2005, Wu and Zhou, 2009, Yeung et al., 2005 and Zhou et al., 2010). Consistently, Leng and Zhou (2010) found that rewarding monetary feedback elicited greater P300, as did observing a friend's, rather than a stranger's, outcomes. While there is evidence that the ACC is related to social psychological constructs like social exclusion, rejection, and unfairness, evidence is only now forthcoming that the FRN component in particular may reflect these constructs. Recently, investigators have used the UG to examine outcome evaluation during decision-making in social contexts. In this social dilemma task, participants typically reject such unfair offers even if the alternative is no monetary gain (Bolton & Zwick, 1995). Unfair offers are characterized by a larger FRN relative to fair offers (e.g., Boksem and De Cremer, 2010 and Polezzi et al., 2008). Boksem and De Cremer (2010) found that the effect was most pronounced in individuals for whom fairness was important. In the context of a social situation, Polezzi et al. (2008) also found that mid-value offers, which could not easily be classified as fair or unfair, also produced larger FRNs relative to fair offers, suggesting that both unfair and mid-value offers may be perceived as undesirable relative to fair offers. The UG and TG, however, do not allow for an understanding of social situations in which mutual cooperation or conflict may occur. The PDG and Chicken Game address this issue, but unlike the well-studied PDG, in which the worst outcome is when the participant cooperates and the competitor aggresses, the Chicken Game has no dominating decision strategy. Because the worst outcome in Chicken Game is when both players compete, there is no clearly advantageous choice. In the present study, participants therefore completed the Chicken Game as their EEG were recorded. Based on the conflict monitoring theory of the FRN, in which outcomes that violate expectancy produce greater FRN amplitudes (Carter et al., 1998, Gehring and Fencsik, 2001, Jia et al., 2007 and van Veen and Carter, 2002), we hypothesized that an individual's FRN amplitude would be greatest when he or she selects cooperation and the other player aggresses. In other words, we expect that the outcome will be perceived as most negative when a player cooperates and the other person violates that trust and aggresses. Similarly, we hypothesized that the attentional and motivational salience of an outcome, reflected in the P300, would be greater if the individual chose to aggress rather than cooperate, as these outcomes were of more monetary risk. We also hypothesized that, consistent with other work in which rewarding outcomes elicit larger P300s (e.g., Leng & Zhou, 2010), the P300 would be larger when the player gained money.

The percentage of trials in which participants selected cooperation and aggression were 44% and 56% (SD: ±7.7%), respectively. Individual decision strategies are presented in Table 2. Paired-samples t-tests indicated that participants selected the aggression strategy significantly more than cooperation, t(14) = −3.202, p = .006. The mean response time for selecting cooperation and aggression were 568.70 ms (SD: 154.06 ms) and 574.26 ms (SD: 162.28 ms), respectively. There was no significant difference between the response times to the two strategies, t(14) = −0.378, p > .05. Table 2. Participants’ decision strategies. On average, participants chose to cooperate 44% of the time and aggress 56% of the time (SD: 7.7%). Participant % Cooperation (C) % Aggression (A) 1 27.0 73.0 2 37.0 63.0 3 46.5 53.5 4 54.0 46.0 5 38.6 61.4 6 40.7 59.3 7 42.5 57.5 8 42.2 57.8 9 59.5 40.5 10 46.7 53.3 11 44.0 56.0 12 50.5 49.5 13 36.5 63.5 14 42.6 57.4 15 45.7 54.3 Table options 3.2. dFRN The main effect of electrode site was significantly quadratic (Fquad(4.84, p < .05, MSE = 2.72, η2 = .257) with roughly equivalent dFRN amplitudes at Fz, FCz and Cz and a diminution of the dFRN at CPz and Pz. While there was a marginally significant main effect of decision strategy, F(1, 14) = 4.35, p = .056, MSE = 39.79, η2 = .237, a significant interaction between decision strategy and electrode on the dFRN amplitude (Flin(1, 14) = 6.32, p < .05, MSE = 5.20, η2 = .311 suggests that the dFRN was largest at Fz and got progressively smaller toward the posterior electrodes when participants chose to cooperate, rather than aggress (see Fig. 2 and Fig. 3). Full-size image (82 K) Fig. 2. Grand average waveforms. Taken across the five electrodes, C indicates cooperation and A indicates aggression. CA indicates that the participant selected cooperation and the opponent selected aggression (−10/+30); CC indicates that the participant and opponent both selected cooperation (+10/+10); AA indicates that the participant and opponent both selected aggression (−30/−30); AC indicates that the participant selected aggression and the opponent selected cooperation (+30/−10). The scalp topography of the dFRN is depicted at 265 ms. Figure options Full-size image (25 K) Fig. 3. Means and standard errors of the amplitude of FRN at Fz (A) and FCz (B). Figure options 3.3. P300 The P300 was significantly larger when participants chose to aggress rather than cooperate, F(1, 14) = 6.92, p = .020, MSE = 62.49, η2 = .331, and when the feedback valence was positive (i.e., when the partner cooperated; F(1, 14) = 4.87, p = .045, MSE = 54.18, η2 = .258; see Fig. 2). The effect of electrode was significantly quadratic, consistent with a centro-parietal distribution of P300, Fquad(1,14) = 27.03, p < .001, MSE = 12.15, η2 = .659. The effect of participant strategy interacted with electrode site, Flin(1,14) = 7.84, p < .05, MSE = 4.09, η2 = .359, such that the difference in P300 amplitude between the aggressive and cooperative strategy diminished toward the posterior midline electrodes. Interactions of decision strategy by feedback valence, and decision strategy by feedback valence by electrode were not significant. 4. Discussion This study investigated the psychophysiological correlates of cooperation and aggression using a task that presented participants with a true social dilemma. With no dominating strategy, the Chicken Game produces an outcome for each player based on the unique combination of their independent decisions to cooperate or compete, much like decisions that individuals face in their daily lives. It is important to note that all participants in the present study believed they were playing a graduate student, which emphasizes the realistic, social nature of the game. By examining players’ evaluations of the four types of outcomes using the dFRN and P300 ERP components, we gained insight into social decision-making processes. The dFRN results depict that, in social decision-making tasks, evaluations of one's own outcomes are heavily influenced by the opponent's decisions. The dFRN was larger when participants chose to cooperate, rather than aggress, particularly at fronto-central sites. As the dFRN is believed to index the extent to which an individual achieves the desired outcome (Holroyd et al., 2006), these findings suggest that an individual may perceive an opponent's decision to aggress most negatively when he or she has chosen to cooperate. In such situations, an individual puts trust in the opponent to also choose the decision strategy that would be of mutual benefit. However, the opponent breaks that trust by choosing to aggress. Although both players choosing to aggress would produce a worse fiscal outcome for the player, the finding that individuals have larger dFRNs when they choose to cooperate suggests that the FRN reflects not merely a loss, but rather the largest loss differential relative to the other player's outcome. Therefore, the dFRN may be sensitive to the trust that one player shares with another. In nonsocial situations, the FRN is typically elicited by negative outcomes, regardless of the magnitude of the outcome (Hajcak, Moser, Holroyd, & Simons, 2006). However, the current data suggest that in a social context, cooperative and competitive tendencies affect how outcomes are evaluated. These data support Sanfey's argument (2003) that the ACC not only takes into account the valence of the outcome, but also emotional responses to unfair offers. It also provides support that the FRN is sensitive to the affective properties of social pain. The P300 demonstrates that people are also motivated both by winning and taking risks. Consistent with previous research (Yeung & Sanfey, 2004) that monetary gains elicit more attention than losses, the participant's P300 was greater when her opponent cooperated, resulting in the participant gaining money. Moreover, there was a significant main effect of decision strategy and a decision strategy by electrode interaction, with greater P300 amplitude following outcomes for which the participant chose to compete, rather than cooperate. The P300 findings suggest that increased attention and motivation are associated with the riskier decision strategy, as choosing to aggress would either cause the participant to gain or lose the most money possible for the given trial. Because P300 amplitude increases as outcomes become less probable (Polich, 1986), it is important to note that in the present study, participants more often chose to cooperate than to aggress and that the subject's choice was followed on half the trials by an aggressive response and on half the trials by a cooperative response. This suggests that the P300 does not merely reflect probability choice or outcome, but because the P300 has a subjective component, it may reflect that more motivation and attention is given to outcomes involving a higher degree of risk. Given the widespread distribution of the P300, the relationship of our findings to cognitive networks and associated neural structures is less certain. P300 generation, in part, may be associated with the ACC, which would corroborate our dFRN results and suggest that the ACC is sensitive to the affective and motivational aspects of the situation. There are several limitations to this study. First, the sample size, although similar to other studies in this field (e.g., Leng & Zhou, 2010), was small. Increasing the sample size may provide more clarity as to the topographical locations of observed effects, particularly regarding the P300, which often is maximal more posteriorally (Polich, 2012). Second, as this study was conducted in China, a question lingers about the generalizability of the findings. Eastern cultures tend to be more collectivistic and may be less motivated toward individual gains than their Western, individualistic counterparts. Comparing effects across cultures or social value orientations would supplement this study to provide a more comprehensive understanding of decision-making. Coupling the Chicken Game with psychophysiological measures allows investigators to objectively tap the cognitive reactions and motivation that shape both every day and large-scale decisions. Specifically, the FRN and P300 help to decipher different aspects of the outcome evaluation step of decision-making. As the FRN indexes an individual's quick assessment as to whether or not the desired outcome was achieved, our results indicate that individuals perceive high conflict situations, reflecting the largest loss differential relative to the opponent, to be most detrimental outcome in social decision-making. The P300 findings add that outcomes have more motivational salience when the player has chosen the more risky, competitive option, as well as when the outcome is in her favor. From here, future research should explore adaptive decision-making to see how subsequent behavior is predicted as a function of psychophysiological responses to previous outcomes, as well as the effects of anticipation and interpersonal relationships on decision-making in situations involving true social dilemmas.