Symmetry of reinforcement in social behavior

A great deal of experimental evidence supports the notion of symmetry with regards to rewarding and aversive outcomes having an equally opposite effect on behavior in non-humans (Schuster & Rachlin, 1968; Farley & Fantino, 1978) as well as in humans (Critchfield et al, 2003). Numerous studies have demonstrated a direct effect of punishment, and empirical data that support conclusions pointing toward a symmetrical law of effect, such that behavior follows the matching law (Herrnstein, 1961) in equally opposing manner when compared to reinforcing outcomes. On the other hand, the domain of Prospect Theory (Kahneman & Tversky, 1979) has found that loss has a disproportionately larger effect on behavior than an equivalent gain, seeming to provide evidence contrary to the symmetry argument. The current experiments were designed to test this argument of symmetry using well- established interactions (1 - the Dictator Game, 2 - the Ultimatum Game, 3 - the Prisoner's Dilemma Game), where the stimuli at stake were tokens with one of two values: a conditioned reward stimulus (time-off a boring task) or a conditioned aversive stimulus (time-in a boring task). Both allocation and cooperative behaviors were measured and correlated significantly with Prosocial, Individualist, and Competitor Social Value Orientation across approx. N=2100 participants. Amount given of reward tokens was equal to the amount kept of aversive tokens; overall both allocation and cooperation data indicate strong support for an argument for empirical symmetry between reinforcers and punishers in social interaction settings. These findings agree with previous research observing symmetry of behavior thus reinforcing the symmetrical law of effect (Thorndike, 1911). The results also indicate the reliability of the social value orientation measure in experiments involving social interaction and division of a utilit