The effect of digital music education on non-cognitive abilities: Empirical evidence from China

Previous research on digital music education has often relied on single-factor perspectives or linear models, which limits the ability to capture complex causal relationships in shaping students' noncognitive outcomes. This study addresses this limitation by adopting a set-theoretic approach that combines Necessary Condition Analysis (NCA) and fuzzy-set Qualitative Comparative Analysis (fsQCA). Drawing on survey data from 329 undergraduates in China, the analysis first shows that no single course design dimension constitutes a necessary condition for high self-esteem (SE) or an internal locus of control (IE). Instead, the results identify multiple equifinal and asymmetric pathways that jointly account for strong noncognitive outcomes. For SE, affective-activation bundles that emphasize instructional attractiveness, interaction design, and community functions are more common. For IE, ability-enabling bundles that stress content quality and resource continuity are more prominent. These findings extend the literature by demonstrating that digital music education should be understood as the outcome of conjunctural causation rather than isolated levers, and they highlight the need for construct-sensitive instructional design. Practically, the study suggests designing for bundles rather than parts, building sustained resource ecosystems, and structuring inclusive interaction to support both self-beliefs and self-regulation. Limitations include the cross-sectional design and reliance on self-reports, which future research can address through longitudinal and behavioral approaches.