Influence of metacognitive prompting on learning within computer-mediated problem sets

Previous studies reported in which students were prompted to use metacognition during problem solving indicated that prompting improves problem solving performance. In this experiment, metacognitive cues were imbedded within computer mediated problem sets (COMPS) and learning was measured. Subjects included two groups of volunteer undergraduates from a course in biological science, with 32 students in a treatment group and 30 students in a comparison group. Treatment group subjects received metacognitive prompts while students in a comparison group completed similar COMPS but without metacognitive cueing. Metacognitive prompts were delivered in two ways: through a faux e-mail program, and by requests to students to write down their thoughts while building an Electronic Notebook. The primary measure of learning, a 34-item posttest indicated significant differences in learning between the two groups. For three quizzes that served as secondary measures of learning there were no differences in mean scores between the groups. Students in the treatment group spent significantly more time completing COMPS but time was not found to be a predictor of learning in this experiment. Qualitative information gathered through tutorials and by interviews showed that students used metacognitive techniques while problem solving. Informal observations indicated that students enjoyed COMPS and were highly motivated to arrive at correct responses to solutions. Based on quantitative and qualitative data this research suggests that including metacognitive training and prompting within hypermedia tutorials may improve students' performance on solving biostatistics problems.