AI Insight
This study examined how self-efficacy relates to English language achievement among Chinese undergraduate students through three types of learning engagement: behavioral, emotional, and cognitive. Using structural equation modeling with 1,000+ students, researchers found that behavioral engagement directly predicted language test scores, while emotional and cognitive engagement served as intermediate steps linking self-efficacy beliefs to behavioral engagement. The sequential model where engagement dimensions build upon each other fit the data better than treating them as separate parallel processes.
Why it matters
The findings suggest that interventions to improve language learning should prioritize fostering actual behavioral engagement (time on task, effort, persistence) rather than focusing solely on emotional or cognitive aspects. Understanding this sequential relationship helps explain why previous studies showed inconsistent results and provides educators with a clearer pathway from building student confidence to achieving measurable learning outcomes.
Understand the Science
BackgroundLearning engagement (LE) has been recognized as a key mediator connecting learning beliefs and achievement, but previous studies have examined its sub-dimensions—behavioral (BE), emotional (EE), and cognitive (CE) engagement—either in isolation or as parallel constructs, overlooking their potential sequential relationships. This analytical approach may obscure how these dimensions work together and yield inconsistent findings.MethodDrawing on Social Cognitive Theory (SCT) and the Broaden-and-Build Theory (BBT), this study employed structural equation modeling (SEM) to compare a sequential mediation model with a parallel alternative among Chinese EFL undergraduates preparing for the CET-4. Self-efficacy and the three engagement dimensions were measured concurrently approximately two months before CET-4, with its scores obtained from official records upon release.ResultsThe sequential model fit the data better than the parallel alternative, consistent with the hypothesized sequential framework. Self-efficacy was associated with BE through independent emotional and cognitive pathways, and a serial pathway with EE theoretically specified prior to CE. Among the engagement dimensions, only BE was directly associated with subsequent L2 achievement. Self-efficacy also retained a direct association with achievement.ConclusionWithin the hypothesized engagement framework, BE was the most proximal engagement dimension associated with internal psychological states and L2 achievement. The findings offer an integrative perspective on past inconsistencies and suggest that self-efficacy is associated with language test performance through both sustained behavioral investment and direct, unmediated processes.