Learning engagement as a protective moderator of the admission score–vocational course achievement association

Introduction Admission scores predict college performance, but engagement may alter this association. We tested learning engagement as a protective moderator and examined related indirect associations. Methods: We matched admission records, a May–June 2026 survey, and same-semester grades extracted July 13, 2026 for 491 Chinese vocational college students; the survey-to-extraction interval was 14–45 days. Results Engagement weakened the admission-score–achievement association ( B = −0.263, p < 0.001; standardized interaction = −0.144, 95% bootstrap CI [−0.214, −0.071]; ΔR 2 = 0.021). The high-engagement slope was imprecise ( B = 0.099, p = 0.078), not demonstrated to be zero; 18.1% of students exceeded the Johnson–Neyman threshold. Self-efficacy showed cross-sectional indirect associations with achievement (ab = 0.321, 95% bootstrap CI [0.232, 0.416]) and career adaptability (ab = 0.153, 95% bootstrap CI [0.106, 0.202]) through engagement. Teacher support modestly moderated the engagement–achievement association ( B = 0.348, p = 0.025; ΔR 2 = 0.009). Results remained stable in integrated and course-adjusted analyses. Discussion Engagement was associated with weaker dependence of achievement on admission scores, but the interactions were small and observational. The indirect pathways and teacher-support interaction require independent replication.