A Questionnaire Method of Class Evaluations Using AHP with a Ternary Graph

Abstract

One of the primary activities for the faculty development of universities is to improve contents and methodology of classes by questionnaire surveys for students. In this study, we propose AHP with a ternary graph to a questionnaire survey for class evaluation in order to measure students’ integrating evaluations which reflect on their concepts of values. Our method has two features. First, a weight vector for three criteria: Expertise, clarity, and personality, is measured not by calculating three combinations of pairwise comparisons but by selecting a grid area in a ternary graph. The proposed method is favorable for anonymous questionnaire surveys because respondents cannot be identified, and thus it is impossible to ask the respondent to reassess the pairwise comparisons if the C.I. value exceeds a threshold. Second, the weighting of the criteria on the goal and that of the alternatives on the criteria are by different two respondents. The former is by the students while the latter is by a faculty member. This division reflects on not only the easiness of assessments but also the consistency with the objective of the class evaluation. In the future work, we can develop a matching system which recommends the suitable classes for students in combination with needs of the students and characteristics of the classes. This is because a student may favor a class favored by the previous students who have similar concept of views on class evaluation.