DEVELOPING TECHNIQUES FOR MEASURING AND ENHANCING STUDENTS' COGNITIVE AND METACOGNITIVE SKILLS

摘要

While previous work established that some self-report measures of metacognition correlate highly with programming and other CS skills [2,3], the current pilot project seeks to develop a behavioral task measure of metacognition and to use exercises modeled on this behavioral task to teach and enhance metacognitive skills. These exercises, based on work by Deanna Kuhn at Columbia University, are to be incorporated in an undergraduate computer science program for the purpose of enhancing cognitive and metacognitive skills that have been identified as important for computer scientists [2,3]. First, the pilot study established a baseline comparison between the MSI [2,3,18] and one of Kuhn 's most widely used causal inferencing tasks [10,11], dubbed the "boat-races" (BOATS) task. Second, the pilot study established a baseline comparison between the BOATS model task and several domain-specific tasks developed by the researchers; these domain specific tasked were modeled after the ones developed by Kuhn. Regression analyses were performed, using the BOATS (domain independent) task, the MSI [2,3] scores, and the MCSS [19] scores as predictor variables and performance on the newly created domain-specific tasks as criterion variables. While none of the individual component variables was found to be predictive of exercise task performance, the overall model was significant for at least two of the three domain-specific tasks. Continued research will use these task scores as predictors and Computer Science task performance as criterion variables. Should the exercises prove predictive, they will he used as training tools to develop metacognitive skill.