Online or Face-to-Face: A Data-Analytics Approach to Understanding First-Year Students' Learning Preferences and Academic Achievements

Abstract

This research explored the predictive power of individual and contextual factors on first-year students' preferences for online versus face-to-face learning environments, examined through a Bayesian framework for analyzing compositional data. Purposefully, it delves into how students' academic performance and geographic location influence their educational modality choices in a post-pandemic context. As quantitative research, the paper employs Bayesian statistical methods, the research analyzed student preferences across varied South African provinces, juxtaposing these with academic performance to uncover patterns and predictors of learning modality preferences. The student learning modalities preferences were collected through questionnaire as a compositional data. The findings revealed a significant correlation analysis which is, that students with higher academic performance and those residing in remote areas show a preference for face-to-face learning modality, challenging the notion that online learning preferences are primarily driven by academic outcomes alone. Adding a novel dimension, this investigation enriches the current understanding of educational preferences by applying a Bayesian approach, revealing that predictive analyses must account for a blend of personal and situational factors. This insight is pivotal for formulating educational policies that are both inclusive and responsive to the diverse needs of the student populace in a dynamically evolving educational terrain.

Keywords

• compositional data analysis
• face-to-face learning
• first-year students
• online learning
• post-pandemic education