A mixed-methods study on the interplay between AI-assisted dynamic assessment, reading comprehension, and...

Dynamic assessment (DA), grounded in Vygotsky's sociocultural theory (SCT), evaluates learners' developmental potential through mediated support within the zone of proximal development (ZPD). However, limited research has examined how AI-assisted DA (AI-DA) may influence both learners' psychological attributes and language performance. This concurrent mixed-methods study investigated the effects of AI-assisted DA (AI-DA) on Chinese English-as-a-Foreign-Language (EFL) learners' academic buoyancy, academic resilience, and reading comprehension. Conducted at a university in China, the study involved two intact classes (N = 92) randomized into an experimental group receiving AI-assisted DA and a control group undergoing teacher-fronted sessions over 16 weeks. Qualitative data were collected through semi-structured interviews and focus group discussions to explore learners' perceptions of buoyancy and resilience, while quantitative data were obtained using a validated reading comprehension test. hematic analysis revealed that AI-DA enhanced learners' academic buoyancy by reducing daily frustrations and strengthened resilience through adaptive scaffolding that encouraged persistence. The quantitative analyses using t-tests and Analysis of Covariance (ANCOVA) demonstrated that the experimental group exhibited significantly higher reading comprehension compared to their peers in the control group. Overall, the findings suggest that AI-assisted dynamic assessment can simultaneously support L2 reading development and learners' psychological adaptability in academic contexts.