Longitudinal clustering of students’ self-regulated learning behaviors in engineering design

• Published in: Computers and education Vol. 153; p. 103899
• Main Authors: Li, Shan, Chen, Guanhua, Xing, Wanli, Zheng, Juan, Xie, Charles
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-08-2020
• Subjects: Engineering design; Learning analytics; Longitudinal clustering; Self-regulated learning; STEM education; Engineering design; Self-regulated learning; Learning analytics; STEM education; Longitudinal clustering
• ISSN: 0360-1315; 1873-782X
• DOI: 10.1016/j.compedu.2020.103899

It is vital to develop an understanding of students' self-regulatory processes in the domains of STEM (Science, Technology, Engineering, and Mathematics) for the quality delivery of STEM education. However, most studies have followed a variable-centered approach, leaving open the question of how specific SRL (Self-regulated Learning) behaviors group within individual learners. Furthermore, little is known about how students' SRL profiles unfold over time in STEM education, specifically in the context of engineering design. In this study, we examined the change of students’ SRL profiles over time as 108 middle school students designed green buildings in a simulation-based computer-aided design (CAD) environment. We identified three distinct SRL profiles using a longitudinal clustering approach: reflective-oriented, adaptive, and minimally self-regulated learners. In addition, we found that students with different profile memberships differed significantly in their design performance. Specifically, adaptive self-regulated learners outperformed minimally self-regulated learners on design completeness, whereas reflective-oriented self-regulated learners demonstrated higher design efficiency than minimally self-regulated learners. This study provides researchers with both theoretical and methodological insights concerning SRL dynamics. Findings from this research also inform practitioners about the design of adaptive interventions.
•We identified three distinct SRL profiles using a longitudinal clustering approach.•This study provides theoretical and methodological insights concerning SRL dynamics.•This study helps improve the delivery of quality STEM education.