This paper investigates how process data like response time and click position relates to economic decisions. We use a social value orientation experiment, which can be considered as a prototypical multi-attribute decision problem. We find that in the social value orientation task more individualistic subjects have shorter response times than prosocial subjects. Individualistic subjects click more often on their own payoffs than on the others’ payoffs, and they click more often on their own payoffs than prosocial subjects. Moreover, the response time information and the click position information are complementary in explaining subjects’ preferences. These results show that response times and click positions can be used as indicators of people’s preferences.
There is ample evidence that people differ considerably in the strength of their preferences. We identify individual heterogeneity in social motives and selfishness in a series of binary three-person dictator games. Based on this identification, we analyze response times in another series of games to investigate the cognitive processes of distributional preferences. We find that the response time increases with the number of conflicts between individually relevant motives and decreases with the utility difference between choice options. The selfish motive is more intuitive for subjects who are more selfish. The heterogeneity in preferences is reflected in the heterogeneity of the underlying cognitive processes. Our findings provide evidence for both, sequential sampling models and dual-process theories, and help to reconcile the mixed results on the correlations between response time and prosociality. Our results also show that it is important to take heterogeneity of preferences into account when investigating the cognitive processes of social decision making.
Cognitive Process of Prosocial and Antisocial Punishment.
This paper investigates the cognitive process underlying prosocial and antisocial punishment using response time and choice data from a public goods game. The results show that response time decreases with the strength of preferences between punishment and non-punishment, which indicates that both the cognitive process of prosocial and antisocial punishment is consistent with the sequential sampling process. In addition, the sequential sampling process starts near the threshold of non-punishment when cooperators punishing defectors, and the process starts near the threshold of punishment when defectors punishing cooperators. That is, prosocial punishment is more deliberative while antisocial punishment is more intuitive than non-punishment. Furthermore, the sequential sampling process of punishment and the intuition or deliberation of punishment can be corroborated by response time difference between punishment and non-punishment decisions and the exercise of the out-of-sample predictions.