Kestin and colleagues · Scientific Reports, 2025
What the study found
In a randomized trial involving 194 university physics students, a purpose designed AI tutor produced higher immediate learning gains than the active learning classroom comparison.
What it does not establish
The trial covered two lessons in one university course. It does not show that ordinary chatbots replace teachers or that the finding applies to every age, subject, or long term outcome.
AN IDEA TO TRY
Make it practical.
Use AI to ask questions, give hints, and help a learner explain their reasoning. Check understanding again without the tool. Follow the school’s rules and use age appropriate tools.
A question to consider: Is the learner doing more of the thinking, or simply getting to an answer faster?
The practical suggestion is an application of the research, not a tested promise about Growthpreneur services.
Back to the resource desk