Necessity and contingency in reasoning (original) (raw)

An experiment was designed to investigate whether subjects perform the conditional falsification problem better on sentences they consider as contingently true than on sentences they consider as necessarily true. Subjects had to generate their own sentences out of which the experimenter selected a necessity and a contingency sentence for the falsification task. Instead of the usual four-card presentation the task was performed by means of an interview which was recorded on tape. Results show that subjects performed significantly better on contingency sentences than on necessity sentences. However, the overall rate of insight was low: eight out of 22 subjects gained complete insight. It is hypothesised that subjects' tendency to look for confirming evidence may be explained as a consequence of cognitive self-reinforcement.