مطالب

Given the transformative potential of technology-enhanced methods in language education, this study investigated the effect of AI-scaffolded learning-oriented assessment (LOA) on Iranian EFL learners' vocabulary acquisition. A quasi-experimental design was employed, with 40 male intermediate-level learners non-randomly assigned to an experimental group (AI-supported Nearpod platform with adaptive scaffolding) and a control group. The Oxford Placement Test ensured homogeneity, while the Vocabulary Knowledge Scale measured vocabulary gains. The experimental group received AI-driven scaffolding, including real-time adaptive feedback, tiered support, and interactive peer-review tasks aligned with LOA principles. Statistical analysis revealed a significant difference in post-VKS scores (M = 130.05) for the experimental group compared to the control group (M = 99.85), with a large effect size (t = 39.38, p < .001, d = 1.96). These results demonstrate that AI-scaffolding, when integrated with LOA, substantially enhances vocabulary learning by personalizing feedback and promoting active engagement. The study highlights the efficacy of AI tools like Nearpod in EFL contexts while underscoring the importance of pedagogical design, such as balancing automation with teacher guidance, to maximize learning outcomes. These findings advocate for the strategic adoption of AI-scaffolded LOA in language curricula, though challenges like digital literacy and equitable access warrant further exploration.