MINDS: metacognitive awareness, identity investment, negative affect, and digital scaffolding as co-predictors of oral English proficiency in undergraduate learners

Introduction Oral English proficiency (OEP) is central to Chinese undergraduates' academic and professional development, yet its cognitive, motivational, affective, and digital determinants are often examined separately. This study tests the MINDS framework, which integrates metacognitive awareness (MA), identity investment (II), negative affect (NA), and digital scaffolding (DS) as co-predictors of OEP. Methods We analyzed an integrated corpus of N = 412 Chinese undergraduate EFL learners constructed from three publicly available datasets: PISA 2022, SpeechOcean762, and a Mendeley IELTS learning-strategies dataset. Structural equation modeling with 5,000 bootstrap resamples tested four direct effects, the mediating role of MA in the II-OEP relationship, and the moderating role of DS in the NA-OEP relationship. Results The measurement model showed acceptable fit (χ −2 (278) = 687.34, CFI = 0.941, TLI = 0.932, RMSEA = 0.059, SRMR = 0.062). MA, II, and DS positively predicted OEP, whereas NA negatively predicted OEP. MA partially mediated the II-OEP association (indirect β = 0.124, 95% CI [0.073, 0.185]), and DS buffered the negative NA-OEP relationship (β int = 0.17, p = 0.001). Together, the four predictors explained 52.1% of OEP variance, and the effect pattern was reproduced across the three datasets. Discussion These findings identify metacognitive strategy support, identity investment, and digitally scaffolded low-stakes speaking practice as complementary leverage points for improving undergraduate EFL oral proficiency.