| The purpose of this investigation was to empirically test a proposed taxonomy of problem solving skills which classified problems according to the cognitive skills required. The research question was whether or not mastery training in a task requiring rule induction had any effect on (1) another task requiring rule induction skills and (2) a task requiring means-end analysis.;A pretest-treatment-posttest design was used in the study. Sixty-nine students enrolled in introductory psychology courses participated. The rule induction tasks were Letter Series Completion Problems and Geometric Analogies. The means-end task was Hobbits and Orcs, a variant of river-crossing problems. The subjects were randomly assigned to one of three groups: two experimental groups who received mastery training in rule induction skills in solving Letter Series Completion Problems and a control group who evaluated mathematical lessons on the PLATO Computer System. All subjects were pretested on Geometric Analogies, Letter Series Completion Problems, and given one attempt to solve the Hobbits and Orcs problem. Following treatment, mastery training or control, all subjects were posttested on the three tasks.;An analysis of covariance indicated that there was no significant impact of the rule induction training on the problem requiring means-end analysis. These tasks appear to require separate skills. An analysis of covariance indicated that there was a significant difference between the experimental groups who received training and the control group on the posttest scores on the Geometric Analogies test, the untrained task which required rule induction skills. Rule induction skills can be taught and are transferable.;These results are an important addition to problem solving research in that rule induction has been shown to be the shared basic cognitive skill involved in two tasks; rule induction skills can be taught; and rule induction skills are transferable. Equally important is the result that rule induction and means-end analysis are separate tasks. A serviceable taxonomy of problem solving skills has been shown to be empirically testable and based on sound psychological theory. |
