Dimensional analysis and predictive models in problem solving

Dimensional analysis is used for the determination of the information‐processing demand (M‐demand) of a problem. A simple predictive model for multi‐step problems is proposed, and its predictions are checked against actual data. It is found that this model is successful in some cases, and unsuccessful in others even in cases of an M‐demand of 2. The effect of various psychological measures of the solvers is examined. A major predictive model for problem solving in science education is the working‐memory overload hypothesis of Johnstone and El‐Banna. This model is based on the working‐memory theory, as well as on Pascual‐Leone's M‐space theory, and predicts that a subject will not be successful in solving a problem, unless the problem has an M‐demand which is less or equal to the subject's M‐capacity. Mechanisms that may block the solution or that may lead to violation of the model are examined, and some necessary conditions for its successful application are discussed in the light of research data.