Development and validation of the intelligent digital learning student engagement influencing factor scale for...

This study aimed to develop and psychometrically evaluate the Intelligent Digital Learning Student Engagement Influencing Factor Scale for nursing and other medical students.

This instrumental development study was conducted in two phases. Phase 1 involved scale development through theoretical modeling, synthesizing the Pedagogy-Social Interaction-Technology Model, the Model of Student Motivation and Self-Regulated Learning, and the Three-Dimensional Model of Student Engagement, followed by qualitative interviews with 25 medical students guided by this integrated framework, a two-round Delphi expert consultation with eight experts, and pilot testing with 30 nursing students. Phase 2 involved psychometric evaluation through a cross-sectional survey of 1,499 medical students, with nursing students constituting the largest subgroup (n = 673, 44.90 %), followed by clinical medicine (n = 616, 41.09 %), medical laboratory science (n = 132, 8.81 %), and preventive medicine (n = 78, 5.20 %). The sample was randomly divided into two subsamples for exploratory (n = 750) and confirmatory (n = 749) factor analyses. Construct validity, reliability, and item discrimination were assessed.

The final scale comprised 20 items across four factors: personal digital literacy, perceived school support, perceived technological environment, and perceived interpersonal interaction and support. These factors validated the integrated theoretical framework, reflecting learner-level psychological mechanisms, macro-level institutional support, and micro-level classroom factors encompassing technological and social interaction dimensions. Exploratory factor analysis explained 66.85 % of the total variance, with factor loadings ranging from 0.488 to 0.751. Confirmatory factor analysis demonstrated acceptable model fit (χ²/df = 4.587, CFI = 0.928, TLI = 0.968, RMSEA = 0.049, SRMR = 0.054). Average variance extracted values ranged from 0.542 to 0.608, and composite reliability ranged from 0.846 to 0.890. The scale demonstrated excellent internal consistency (Cronbach's α = 0.920) and split-half reliability (0.918). Item discrimination analysis confirmed significant differences between the upper and lower 27 % of the sample (P < 0.01).

This scale enables nurse educators to diagnose engagement deficits across institutional, technological, and interpersonal dimensions, guiding curricular improvements and providing a basis for precise interventions amid digital teaching transformation. Future studies should longitudinally examine its utility in nursing populations and its predictive validity for clinical performance.