Previous studies on the representative role compare representatives’ behaviors with individuals’ behaviors. Nevertheless, since representatives are a particular type of group members, previous designs inevitably did not distinguish between in-group favoritism effects and the actual effect of the representative role. Carrying on this methodological criticism, the present thesis suggests a new experimental design aimed at disentangling the two effects. Specifically, this study’s aim is to causally determine the effect of the representative role on individuals’ cooperation and understand which mechanisms foster the representation effect. To do so, an online experiment was conducted on Amazon Mechanical Turk. The results show that neither the in-group favoritism effect nor the effect of the representative role alone seem to have any significant influence on participants’ cooperative behaviors, but by conducting the same analysis done by previous studies a decrease in cooperation for representatives can be detected. However, previous interpretations are criticized and the decrease in cooperation is attributed to the interaction between the representation effect and the in-group favoritism effect and not to the representation effect alone. Additionally, self-reported data collected in the post-experimental survey displays that representatives do feel accountable for the represented group mates and, therefore, are willing to equally share their payoff. Finally, for future studies, this research suggests the use of more efficient minimal group procedures to better understand the isolated representation effect and the use of actual behaviors to investigate the mechanisms that lie at the basis of the representative role.