A Statistical and Optimization-Based Framework for Evaluating Digital Learning Platforms Using Student Survey Data

The increasing adoption of digital learning platforms in higher education has led to the widespread use of student survey data to assess learning experiences and outcomes. While such data provide valuable insights into learners’ perceptions, their analytical potential is often underexploited due to the predominance of descriptive approaches. This paper proposes a statistical and optimization-based framework for evaluating digital learning platforms using student survey data. Perceived learning improvement is modeled as an ordinal response variable influenced by multiple explanatory factors, including accessibility, motivation, content adequacy, technical constraints, time availability, and external practice. The empirical analysis relies on survey data collected from 300 undergraduate students and combines multivariate regression modeling with exploratory factor analysis to identify key determinants and latent dimensions underlying student perceptions. The statistical results are subsequently embedded within an optimization framework that formalizes learning effectiveness as an objective function under practical resource constraints. The findings reveal that motivation and content adequacy are the most influential factors in explaining perceived learning improvement, while technical constraints exert a negative but secondary effect. By integrating statistical inference with optimization-oriented reasoning, the proposed framework provides a structured and decision-relevant approach for the quantitative evaluation of digital learning platforms in higher education.