University Students’ Academic Behaviors by Online Learning Clusters: Focusing on Self-Regulated Learning and Basic Psychological Needs Satisfaction

The rise of online learning environments in higher education has highlighted the need to examine such environments and to explore the psychological characteristics of learners engaged in online learning. In response, this study aimed to identify types (i.e., clusters) of college students based on self-regulated learning and the satisfaction of basic psychological needs, and to analyze differences in academic behaviors across the clusters. A total of 383 undergraduate students (165 female; Mage = 22.4) from four-year universities in South Korea, all of whom had taken at least one semester of online courses, participated in an online survey. A cluster analysis was conducted based on sub-factors of self-regulated learning (metacognition, help-seeking, persistence, environmental structuring) and the satisfaction of basic psychological needs (autonomy, competence, relatedness). As a result, four learner clusters were identified: High (n = 99, 26%), Moderate-High with Low Relationship (n = 84, 22%), Moderate (n = 120, 31%), and Low (n = 84, 22%) clusters. To examine differences in academic behaviors across these clusters, perceived academic achievement, intention to persist in online courses, and academic procrastination were used as criterion variables, while digital device usage was controlled. MANOVA and MANCOVA analyses were conducted to compare cluster differences. The results showed that the High, Moderate-High with Low Relationship, Moderate, and Low clusters exhibited progressively lower levels of perceived academic achievement and course persistence, and increasingly higher levels of academic procrastination, in that order. These findings suggest that higher levels of self-regulated learning and basic psychological need satisfaction are associated with more adaptive responses to online learning environments. Notably, the identification of the Moderate-High with Low Relationship through cluster analysis suggests that low relatedness may be a defining characteristic of certain online learners—a distinction that may not be readily detected through variable-oriented approaches. This finding carries important educational implications, indicating that providing emotional support for students' relatedness in online environments may enhance their adaptation and engagement in such settings.