2026 ASEE Annual Conference & Exposition

This Complete Research Paper details a study on the socio-psychological factors that sustain academic commitment—and, by extension, persistence—in first-year engineering students. The first-year transition is a period of heightened uncertainty during which cognitive and emotional resources significantly influence day-to-day engagement. Building on Social Cognitive Theory, Self-Determination Theory, Expectancy-Value Theory, and models of academic and social integration, we investigate how three personal factors—self-efficacy, self-regulation, and intrinsic motivation—translate into academic commitment through two proximal mechanisms: sense of belonging to the significant and perceived value of coursework. We conducted a cross-sectional study in a required first-semester engineering course (N = 554) at a large engineering school in Chile. Students completed a validated Likert-scale instrument measuring self-efficacy, self-regulation, intrinsic motivation, sense of belonging to the major, perceived value of coursework, and academic commitment (behavioral/cognitive engagement). We estimated multiple-mediator path models using observed variables, employing robust standard errors and confidence intervals. Three theoretically motivated specifications were compared: parallel mediation via belonging and value (Model 1); single mediator belonging (Model 2); and single mediator value (Model 3). Model adequacy was evaluated with robust CFI/TLI, SRMR, and information criteria (AIC/BIC). We report explained variance (R²) and standardized effects. The two-mediator model (Model 1) provided the best fit (robust CFI ≈ .96; SRMR ≈ .03; lowest AIC/BIC) and explained 41% of the variance in academic commitment (R² = .409). Both mediators were well explained by the predictors (R²belonging = .306; R²value = .291) and exhibited significant residual covariance, consistent with their conceptual proximity. In Model 1, self-efficacy showed partial mediation, combining a significant direct effect on commitment (β ≈ .275, p < .001) with a total indirect effect (β ≈ .190, p < .001) through belonging and value. In contrast, intrinsic motivation (β_indirect ≈ .126, p < .001) and self-regulation (β_indirect ≈ .145, p < .001) operated primarily through the mediators; their direct paths to commitment were not statistically significant once mediators were included. Among the mediators, sense of belonging exerted the stronger proximal effect on commitment (β ≈ .313, p < .001) relative to perceived value (β ≈ .169, p < .001). Sensitivity analyses showed that models with only one mediator performed less effectively. Removing value (Model 2) decreased the model's fit and simplicity, while excluding belonging (Model 3) caused the greatest decline, reducing R² for commitment to .355. Collectively, the evidence supports complementary pathways through which personal resources translate into day-to-day engagement: identity/connection (belonging) and purpose/relevance (value). The findings suggest efforts to improve persistence should target immediate mechanisms. First, fostering belonging—helping students see themselves as part of the engineering community and connected to peers and faculty—can significantly improve commitment during the first year when “belonging uncertainty” is high. Second, course and program designs that highlight and reinforce the value by explaining why concepts and tasks matter for students’ careers and societal impact add explanatory power beyond belonging alone. Third, because self-efficacy links directly to commitment, programs should offer experiences that help students recognize progress and build confidence early and consistently. Fourth, self-regulation is an enabling skill that boosts belonging and value; supporting goal setting, planning, and monitoring helps sustain engagement. Addressing these factors during the first year is vital, as identities are shaped, expectations set, and habits developed. Tackling belonging, value, self-efficacy, and self-regulation together creates a framework to align instructional practices, supports, and advising with the psychological factors most influential based on current evidence commitment. This study explains how core socio-psychological factors—belonging and perceived value—drive academic commitment, with self-efficacy providing additional support. The model guides the alignment of first-year practices and supports to enhance commitment and retention. Limitations include its cross-sectional design and reliance on observed scores, which may compromise the results. Future research will test longitudinal models and subgroup variations. Overall, findings guide first-year engineering efforts to boost engagement, persistence, and success.