Defining predictors of student satisfaction based on student evaluation of teaching using decision tree analysis

Abstract

Student evaluation of teaching is a prevalent method to assess instructional quality and student satisfaction in higher education all over the world. However, there is an ongoing debate as to which characteristics of instructors make them effective. This study aimed to discover which instructional characteristics can predict student satisfaction levels. To this end, a CHAID analysis, a form of decision tree analysis, was conducted on SPSS to reveal the relationships between instructional characteristics of instructors and student satisfaction level measured by a SET form. The study was conducted at an English language preparatory school of a non-profit private university in Türkiye. 4281 forms including 23 Likert-type questions were analyzed. The findings show effectiveness and being supportive are the most significant predictors of student satisfaction. Following them, enabling students to evaluate different perspectives, encouragement to share views, feedback, and positivity are highly valued by students. Less significant predictors are found to be variety of activities, asking questions to encourage students to express opinions, subject knowledge, guidance, encouraging active participation, preparedness, and recommending publications in English. All in all, 13 of 23 items were significant predictors of student satisfaction.