The Effects of Flipped Classrooms in Higher Education: A Causal Machine Learning Analysis

2025-07-14

AI Summary

This study investigates the causal impact of transitioning from hybrid lecture-based instruction to a flipped classroom model in undergraduate statistics courses. Employing double/debiased machine learning (DML) as the primary identification strategy—complemented by propensity score matching and panel data analysis—we address both selection and implementation biases using fine-grained learning behavioral data (e.g., pre-class video engagement, sequential quiz responses). Results indicate that the flipped classroom significantly enhances students’ metacognitive awareness and reduces procrastination, yet yields no improvement in final exam performance or long-term knowledge retention; moreover, perceived enjoyment declines, primarily due to low compliance with pre-class learning tasks. The key contribution lies in the first systematic application of causal machine learning to attribute flipped classroom outcomes, uncovering a critical “intention–execution gap” between pedagogical design and actual implementation. This advances a methodological paradigm for precise, evidence-based evaluation and iterative optimization of educational interventions.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Machine learning and data science for the Web

Abstract

This study uses double/debiased machine learning to evaluate the impact of transitioning from lecture-based blended teaching to a flipped classroom concept in a cohort comparison of a large compulsory introductory statistics course at a German tuition-free university. Our findings indicate positive changes in students' self-conception and a reduction in procrastination behaviors. However, we also observe a decline in the enjoyment of classroom sessions. Contrary to theoretical expectations, we do not find significant positive effects on exam scores, passing rates, or knowledge retention. Unlike most studies, however, we can leverage detailed usage data from the flipped cohort, including the timeliness and completeness of pre-class video watching, as well as quiz participation patterns, to check how well students implemented each part of the curriculum. Our findings suggest that, on average, students in the flipped cohort implemented the instructional approach insufficiently, explaining the mechanism of our null results in exam performance and knowledge retention. This highlights the need for additional strategies to ensure that students actually benefit from a flipped curriculum.

Problem

• Evaluates flipped classroom impact on student outcomes
• Analyzes student engagement with pre-class materials
• Identifies insufficient implementation of flipped curriculum

Innovation

• Uses double/debiased machine learning for causal analysis
• Leverages detailed usage data from flipped cohort
• Analyzes timeliness and completeness of video watching