Hypothesis testing in Wason's selection task: Social exchange cheating detection or task understanding

Recently, an evolutionary view of performance in the Wason selection task was proposed, according to which people successfully solve tasks involving social exchange situations, or cheating detection content and perspective, but fail to do so in other domains. Alternatively, we propose that performance in the Wason problem largely depends on three aspects related to how people understand the task: (1) the clarity of the rule in terms of determination and direction; (2) the nature of the alternative to the tested rule and the falsifying instance it entails; (3) the perceived relevance of looking for violation strategy. We show that Gigerenzer and Hug's improvement in performance with "cheating" compared to "no-cheating" versions can be explained by these elements of task understanding rather than by cheating. In Study 1 facilitative understanding features were removed from the cheating versions and were introduced into the no-cheating versions, without affecting the cheating (or the no-cheating) nature of the task or changing perspective. Performance levels in the original cheating and the unconfounded no-cheating versions were found to be equally high (71%), whereas the unconfounded cheating and the original no-cheating versions yielded equally low performance (30-32%). Study 2 showed that the reversal in choice patterns obtained by Gigerenzer and Hug by changing perspectives in bilateral cheating option rules can be achieved without changing perspectives. Moreover, this reversal fails to occur when perspective change does not accompany change in task understanding.