Effects of self-directed learning model based on AI-driven adaptive learning system on learning achievement of teacher qualification examination

Abstract

University students increasingly pursued professional qualifications such as the Teacher Qualification Examination (TQE) to improve employability. However, they faced challenges in self-directed learning (SDL) due to difficulties in monitoring progress and identifying resources. This study aimed to examine the effects of an SDL model based on an Artificial Intelligence (AI)-driven Adaptive Learning System (ALS) on TQE learning achievement among students majoring outside of teaching fields. The research employed an experimental design with 80 TQE candidates selected through simple random sampling. Validated instruments included SDL-ALS lesson plans and a 50-item Educational Teaching Knowledge and Competence Test. Data analysis was conducted using Repeated-measures ANOVA. The results indicated significant achievement gains from pre-test (M = 51.40, SD = 13.44) to post-test (M = 92.10, SD = 11.07) and retention-test (M = 92.83, SD = 10.82) (p < 0.001, η² = 0.988). There was no significant difference between the post-test and retention test (p > 0.05). Additionally, lower scores in the SD retention test suggested greater gains among students with initially lower performance. The findings concluded that the SDL-ALS model significantly improved TQE knowledge mastery and retention. Its design effectively addressed individual learning gaps through adaptive scaffolding and structured self-direction. This equity-enhancing SDL-ALS model provided undergraduate students with an effective self-directed learning tool, transformed teachers into diagnostic mentors, and offered researchers a validated framework for integrating adaptive AI with structured pedagogy.