TRAINING IN INFORMATION PROCESSING ALGORITHMS
Michael J. Lawson, John R. Kirby · British Journal of Educational Psychology · 1981
S ummary . Following Hunt's (1975) analysis of the Raven's Progressive Matrices (RPM) test, an attempt was made to identify and train two strategies for solving RPM problems. Four training groups were formed: one received Gestalt strategy training, one received Analytic strategy training, one received training in both strategies, and a control group received neutral instructions. Following training, subjects completed a set of ambiguous items and Set I of the Advanced Progressive Matrices. In the latter, they were required to justify their responses. Evidence is presented which suggests that the two strategies proposed in Hunt's analysis can be identified and that subjects can be trained to use and maintain these strategies. Strategy use is also related to post‐test performance. Analysis of subjects' justifications for responses indicates that Hunt's item analysis of certain RPM problems needs to be modified to take into account both the influence of item‐type and subjects' flexibility in switching between strategies even after training.