Do Gains from Generative AI-Enabled Adaptive Pretesting Persist? Evidence from a Retention Study

Akgun and Toker (2026) examine whether learning gains from GenAI-enabled adaptive pretesting persist over a seven-week retention period. Undergraduate participants completed adaptive AI-assisted pretesting, received instruction, took a baseline assessment, and were randomly assigned to three follow-up conditions: adaptive spaced retrieval practice, fixed spaced retrieval practice, or learner-directed AI-supported study. Multivariate analyses showed significant effects of condition on posttest performance and practice effort, with both retrieval-based conditions significantly outperforming learner-directed study. The findings indicate that while AI-adaptive pretesting can elevate initial understanding — especially for higher-order reasoning — sustained learning critically depends on how subsequent AI-supported practice is structured. This RCT contributes to the Adaptive Learning literature by showing that initial gains are fragile without structured retrieval, and has practical implications for designing Personalized Learning sequences that maximize learning-gains and promote Transfer Of Learning.

Connected Concepts

Connected Articles

Citation

Mahir Akgun, Sacip Toker (2026). Do Gains from Generative AI-Enabled Adaptive Pretesting Persist? Evidence from a Retention Study. arXiv:2606.22328. 27th International Conference on AI in Education