The effects of problem content and negation on Wason's selection task

The present study examined performance on Wason's selection task for two types of thematic content that have been shown to lead reliably to correct responding in this task. The four versions of the implication rule that are possible when negative components (antecedent and consequent) are allowed were used. This permitted a test of the hypothesis that matching bias and verification bias are cognitive short-circuiting strategies which are used when subjects have no prior experience with the problem content. In support of this hypothesis, facilitation was observed for the negated thematic rules when compared to performance on negated abstract rules. In addition, neither matching bias nor verification bias explained very well the data for negated abstract rules which were recodable into non-negative form. For both thematic and abstract problems, a fairly substantial percentage of the selections were unaccounted for with respect to processing strategy.