Hypothesis recognition failure in conjunctive and disjunctive concept-identification tasks

Abstract

While solving either a conjunctive or a disjunctive concept-identification problem, college students were asked to recognize hypothesis, stimulus, and feedback information from the immediately preceding trial. After this phase of the experiment, subjects were asked to estimate feature frequencies of occurrence and to classify old and new instances of the concept. Recognition performance was best for feedback and worst for stimulus information. Contrary to hypothesis theory, hypotheses were correctly recognized 65% of the time overall. Instances presented twice, once, and never before were classified equally well, a finding that argues against specific-instance theory. Finally, frequency estimates increased as a function of actual frequency. All obtained results support both frequency theory and dual-process theory.