CONTEXTUAL FACTORS, STRATEGIC KNOWLEDGE, AND CAPACITY LIMITS IN CONTROL-OF-VARIABLES TASKS

In two experiments we examined the influence of different task conditi ons on the control-of-variables ability. In the first experiment, 202 students, ranging in age from 11 to 14 years, solved a series of paper and pencil tasks demanding the application of the control-of-variable s strategy. In these tasks the students had to develop a solution plan . The results indicated that the solution rate was influenced by the s tudents' familiarity with the context of the task and the distinctiven ess of the variables. The number of variables to be controlled had no effect on the solution rate. In order to get further information, we c onducted a second experiment. 88 students, who had participated in the first study, had to solve two control-of-variables tasks that demande d the actual manipulation of the variables. In contrast to the first s tudy, the results showed an effect of the number of variables to be co ntrolled. That is, the solution rates decreased with increasing number of variables when the students could not use the environment of the t ask as an external store. It seems that the increased number of variab les put a higher load on working memory. In summary, rather than indic ating a formal logical network, our results indicate that the performa nce in formal operational tasks is best described by models assuming t he availability of pragmatic reasoning schemes, domain-specific knowle dge and the necessary capacity of working memory.