Summary of https://bera-journals.onlinelibrary.wiley.com/doi/10.1111/bjet.13544
Investigates the effects of using generative AI, specifically ChatGPT, on university students' learning. A randomized controlled trial compared students using ChatGPT to those using human expert support, writing analytics tools, or no support at all.
The study examined intrinsic motivation, self-regulated learning processes, and learning performance across different groups. Results showed ChatGPT improved essay scores but did not significantly enhance motivation or knowledge transfer, raising concerns about "metacognitive laziness"—over-reliance on AI hindering deeper learning.
The study concludes that AI should supplement, not replace, human interaction in education.