Proceedings of the Annual Meeting of the Cognitive Science Society

Co-operation and group-serving behaviour of group members has increasingly been acknowledged as essential to theflourishing of groups in general and the success of teams in organizations or companies in particular. Studying this,however, presupposes dissociating individual-level and group-level effects (involving a Simpsons Paradox). We havestarted investigating settings where true individual- and group-level effects could be dissociated in a learning paradigmconcerned with individuals in changing teams. Our results show that participants often evaluated the overall most effectivegroup-serving team-player much more negatively than all less effective non-interacting workers. This suggested a potentialTragedy of Personnel Selection, when personnel managers, relying on number-based outcomes, tend to ignore even strongand crucial group-level effects of team-players. Here we briefly summarize some findings and present an experiment,where we tried to improve participants ability to dissociate individual- from group-level effects, by explicitly providingthem with hypotheses about a team player.