Subjective Reasons of Selections in the Wason Selection Task

Abstract

The Wason selection tasks of both abstract standard version and its negated variant, i. e., if “p”then not “q”, were given to 36 subjects, who were then asked to decide whether each card should be checked or not and make a confidence rating between 1 and 5. They were also asked to provide retrospective protocols after both tasks, by being questioned about the reasons for each decision. When they showed inconsistency in their reasons, further prompting questions were made till it was resolved in their mind. The selection data showed that the subjects were more likely to select cards of which the contents were explicit in conditionals, supporting a matching bias account (Evans & Lynch, 1973), but from the protocol data it could be inferred that in an affirmative standard version, some subjects made confirmation while others made biconditional interpretations after an intuitive relevance judgment.