Development and Validation of a Belonging-Supportive Deep Learning Environment Instrument for Evaluating Flipped Learning Environments in Higher Education

Abstract

Supportive learning environments are central to student engagement, psychological well-being, and academic success in higher education. In flipped learning, however, pedagogical effectiveness depends not only on pre-class exposure and in-class active learning but also on whether students perceive the learning environment as socially inclusive, academically supportive, and conducive to deep learning. This study developed and validated the Belonging-Supportive Deep Learning Environment (BSDLE) instrument as an indicator of learning-environment effectiveness in higher education. The study used an instrument-development design with a quantitative validation phase involving 1,161 students from five Indonesian higher education institutions that had implemented flipped learning. Instrument development involved literature review, construct mapping, item generation, expert judgment, and psychometric evaluation. Data were analyzed using a multidimensional Rasch framework. Model comparison showed that the Partial Credit Model (PCM) provided a better fit than the Rating Scale Model (RSM), as indicated by lower final deviance, Akaike Information Criterion, and Bayesian Information Criterion values. Item difficulty ranged from -0.79 to 0.72 logits, and all item fit indices were within the acceptable mean-square range of 0.5 to 1.5. Five BSDLE dimensions demonstrated high EAP/PV reliability (> .90), whereas the faculty-relations dimension showed lower reliability (.695), indicating the need for further refinement. Rating-scale diagnostics suggested that a four-category response format was generally more stable than a five-category format, although several items required revision because of non-monotonic category functioning. Overall, the BSDLE instrument provides promising validity evidence for evaluating belonging-supportive and deep-learning-oriented environments in flipped higher education, while further testing across broader institutional contexts is recommended.