Higher-level causal reasoning for diagnosing students' misconceptions in learning semiconductor physics

A major feature of intelligent tutoring systems is the analysis of a student's problem solving approach and identification of his/her misconceptions about a subject matter. Intelligent tutoring systems generally offer diagnostic tests to observe the student's problem solving methodology and then analyze the solution/(s) to generate remedial instructions. Employing causal reasoning at first-principal level usually performs the diagnosis of students' misconceptions. Causal reasoning at first-principal level is flexible and results in a robust system but is tedious, results in a larger and complex computer code, and is not necessary in all cases. We investigated the use of higher-level causal reasoning in diagnosing a students misconceptions in the area of Semiconductor Physics. We are in the process of revising an intelligent tutoring system for teaching Semiconductor Physics. This paper presents an on-going project of incorporating higher-level causal reasoning for diagnosing students misconceptions and generating remedial instruction set/(s) based upon the diagnosis.