| Abstract: | The study described in this article examined the factors in the development of reasoning on the balance beam. The results of this and other investigations by the author led to the development of a model of problem solving and of the transition between levels of performance, that is, from novice to expert behaviors in a specific problem domain. The sample for this study consisted of 34 nonscience majors enrolled in Physical Science I at the University of Southern Mississippi. The results of this study showed (a) significant correlations between the predictor variables of numerical inductive reasoning ability and ratio span and the dependent measure of amount of practice until mastery of balance beam problems, (b) that there existed significant differences on the predictors between subjects who used the ratio scheme versus those who used the product-moment rule to do balance beam problems, (c) that the sequence of behaviors until mastery of the balance beam was not invariant, and (d) that there existed significant negative correlations between individuals' tendencies to avoid unsuccessful algorithms and the frequency of incidences of monitoring the learning process on one side and the amount of practice and the construct of short-term storage space on the other side. The results of this and other studies by the author are used to develop a model of problem solving and of the development of problem-solving behaviors. This model also includes recent findings in short-term memory and problem-solving research. The implications of this model for instruction are discussed. |
