Social learning strategies, success bias, and similarity.
We are developing an experimental design to uncover heterogeneity in social learning strategies. Social learning strategies are at the core of cultural evolution models. The problem is that we do not know how complex these learning strategies are and small variances in complexity would have significant effects in the models' outputs. Thus better understanding of these learning mechanisms is needed. In this study, we investigate the impact of success bias, similarity, and level of information on social learning strategies. Participants are confronted to 3 different types of social information: a successful reported behavior, the group affiliation of the successful demonstrator (similarity), and the value of the group affiliation information. We manipulate the second information (i.e. in-group or out-group) and the third information (i.e. information reliability). Our goal is twofold. First, to estimate the use of each type of information by generating the general function between the observed behavior and the social learner's behavior. Second, to analyze the diversity of social learning strategies at the individual level.