Abstract:
Stereotypical assumptions made during insight problemsolving can trigger an initial representation that constrains problem solving (Ohlsson, 1992). Two experiments were conducted to investigate the effect of such assumptions on solution rates and hypothesis generation using verbal insight problems, which have received little attention in the literature. Concurrent verbal protocols were collected in both experiments. In Experiment 1, twelve participants attempted two verbal insight problems: the Unseen Walker and Bombs Away. Solution rates for both problems were poor. Qualitative analysis of verbalizations revealed that participants generated incorrect hypotheses on the basis of an incorrect stereotypical assumption. Experiment 2 aimed to improve performance on the same verbal insight problems through generic training to identify inconsistencies between the problem solver’s representation and the problem specification. After training, twelve participants completed the test problems (T condition) and another twelve were permitted to use an aide memoire in order to reduce cognitive load (TA condition: Pfeiffer, 2004). Experiment 1 served as a no training (NT) control condition. Training improved solution rates although the pattern of results varied between problems, such that the aide memoire was more beneficial for the Bombs Away problem. Results supported the notion that stereotypical assumptions can inhibit solution of verbal insight problems and that fine-grained training involving inconsistency checking between the problem statement and a person’s interpretation of it has some benefit at overcoming these barriers.