Network Analysis of Students' Use of Representations in Problem Solving

We present the preliminary results of a study on student use of representations in problem solving within the Modeling Instruction – Electricity and Magnetism (MI-E\&M) course. Representational competence is a critical skill needed for students to develop a sophisticated understanding of college science topics and to succeed in their science courses. In this study, 70 students from the MI-E\&M, calculus-based course were given a survey of 25 physics problem statements both pre- and post- instruction, covering both Newtonian Mechanics and Electricity and Magnetism (E\&M). For each problem statement, students were asked which representations they would use in that given situation. We analyze the survey results through network analysis, identifying which representations are linked together in which contexts. We also compare the representation networks for those students who had already taken the first-semester Modeling Instruction Mechanics course and those students who had taken a non-Modeling Mechanics course.