An analysis of course evaluation questionnaire by machine learning

Course evaluation by questionnaire is the most popular assessment tool for the faculty development of University in Japan. However, many universities do not fully utilize the result of these questionnaires. Usually, traditional statistical tools such as average or standard deviation are used to analyze the questionnaires. Recently, more advanced statistical tools are available and many of them come from the field of machine learning. We thus tried to analyze the course evaluation questionnaire using the advanced statistical tools. In this paper, the questionnaires obtained at National Defense Academy of Japan were analyzed. This questionnaire has been conducted for 10 years history to more than thousand cadets, but the questionnaires have not been analyzed up to the present. This study analyzed and evaluated the reliability of the class questionnaire through the statistical analysis. As a result, potential factors and their relationship were discovered in a questionnaire using Structural Equation Modeling.