What Are the Patterns Between Personality and STEM Attitudes? A Pilot Study With Educational Data Mining Approach

Conference: The Paris Conference on Education (PCE2022)
Title: What Are the Patterns Between Personality and STEM Attitudes? A Pilot Study With Educational Data Mining Approach
Stream: Design, Implementation & Assessment of Innovative Technologies in Education
Presentation Type: Oral Presentation
Ülke Balci Yeşilkaya, Uludağ University, Turkey
İlker Yeşilkaya, University of Balıkesir, Turkey
Salih Çepni, Uludağ University, Turkey
Salih Tutun, Washington University in St. Louis, United States


This research is a pilot study that examines the relationships between vocational preferences and personality types of secondary school students in terms of their attitudes towards STEM by using data mining techniques. Educational Data mining is an important approach in terms of detecting hidden patterns among data and modeling student behaviors. The sample of the study consists of 97 secondary school students in Turkey. In the research, Holland personality types based personal sphere inventory short form and attitude scale were used as data collection tools. Misleading answers were removed with the Fast Minimum Distance Covariance method. Similarities among students were calculated using the Heteregenous function. Two different adjencency matrices were created according to the STEM and RAISEC scale scores. The intersection values of the students in the adjencency matrix define the relationships between them. These relationships form the edges between nodes. Each student is defined as a node. By using these nodes and edges, a graph theory-based network topology is created. In this network, the communities to which the students belong were determined using the Luvian Algorithm. As a result, students representing the communities within the models were determined. The values of these students' STEM attitudes and RAISEC Personality types were compared. Significant patterns were observed in individuals within the communities. The findings suggest that comprehensive studies are needed to better explain these patterns.

Conference Comments & Feedback

Place a comment using your LinkedIn profile


Share on activity feed

Powered by WP LinkPress

Share this Presentation