Not All Active Learning Is Equal: Cluster Analysis of Instructional Talk Patterns and Student Engagement in Undergraduate Statistics Courses

• In Event: Innovative approaches to improving postsecondary teaching

Abstract

This study examines instructional discourse patterns in undergraduate statistics courses using audio analytics from TeachFX. Cluster analysis of six semesters of classroom talk data identified three instructional profiles: individual active learning, collaborative active learning, and lecture interspersed with peer interaction. Linking these clusters to student survey responses and courseware usage, we find that classes combining lecture and collaborative activities are associated with significantly higher affective engagement, instructor belonging, and digital tool usage than those emphasizing individual active learning. No significant differences in assessment scores were observed. These findings challenge the assumption that all active learning is equally effective and underscore the importance of instructional structure. Implications for pedagogy, learning analytics, and faculty development are discussed.

Authors

• Julie Neisler, Digital Promise

• Xin Wei, Digital Promise

• Taylor Alexander, Digital Promise

• Barbara M. Means, Digital Promise