Hubert Feger, Ulrich von Hecker

• Flament and Apfelbaum observed that differences in resource possession lead to group structure in the form of exchange coalitions and hierarchical structure in sociometric choices. We showed that these effects occur even with a uniform distribution of resources. However, knowing the initial distribution of the resources is useful for predicting the final distribution of positive messages received during the expreriment. the two tests we performed indicate that participants in the early part ofthe experiment either react at random or intentionally create a uniform distribution of the positive messages they send to the other positions. Ordered exchange behavior depends on a certain minimum amount of information about the behavior of the others being available to (almost) all participants. We report formal analyses of the reciprocity and the debtor principles. Row reciprocity assumes the return of positive messages to senders without further consideration. Under optimal conditions, it leads within few trials to fixed exchange coalitions.Flament and Apfelbaum observed that differences in resource possession lead to group structure in the form of exchange coalitions and hierarchical structure in sociometric choices. We showed that these effects occur even with a uniform distribution of resources. However, knowing the initial distribution of the resources is useful for predicting the final distribution of positive messages received during the expreriment. the two tests we performed indicate that participants in the early part ofthe experiment either react at random or intentionally create a uniform distribution of the positive messages they send to the other positions. Ordered exchange behavior depends on a certain minimum amount of information about the behavior of the others being available to (almost) all participants. We report formal analyses of the reciprocity and the debtor principles. Row reciprocity assumes the return of positive messages to senders without further consideration. Under optimal conditions, it leads within few trials to fixed exchange coalitions. Relative reciprocity uses the information on the amount of positive messages received from the other participants as well.…