Developing a Fuzzy Clustering-Based Method for Categorizing Young Adolescent Students Based on Their Empathy Scores and Exploring the Relationship Between Their Empathy and Learning Behaviors

Abstract

Background: Empathy is a skill that has been proved effective in learning and teaching processes. Objectives: The aim of this study was to explore the relationship between students’ empathy and their learning behaviors. Methods: A fuzzy clustering-based method (an area of artificial intelligence) was used, according to which students were classified to clusters based on their empathy measures. Students’ empathy was assessed through a questionnaire. Overall, 345 students (11 to 13 years old) from six schools located in three different areas of Tehran, Iran, participated in this study, selected by multistage cluster sampling. In this method, similar samples are classified in one cluster and, then, clusters can be labeled based on their attributes (empathy measures). Two teacher-reported and student-reported questionnaires were used to assess the learning behavior and empathy levels of students. Questionnaires were completed by the students and their teachers during school year 2017 and 2018 (from autumn 2017 to spring 2018). All calculations were performed in MATLAB, a multi-purpose programming environment. Results: Although statistical parameters showed a strong relationship between students’ empathy and their learning behaviors, AI clustering process provides a more exact analysis due to its nature. The results revealed a significant relationship between empathy scores and learning among male students. A P value of 0.0031 indicates a meaningful relationship between empathy scores and learning behavior measures. Conclusions: Number of students in each cluster showed that females are more uniform than males in the sense of empathy. Cultural backgrounds have significant effects on answers to questions. Processes revealed a meaningful difference between males and females when their connection of empathy and learning behaviors were investigated. Cognitive components seem to be more de-terminative than affective components.