AI tutoring matches human effectiveness for GRE preparation, according to a new study.

AI tutoring matches human effectiveness for GRE preparation, according to a new study. Researchers found that AI tutors produced learning gains statistically equivalent to those of expert human tutors.

A study utilizing the StudentBench platform found that artificial intelligence tutors provided learning gains for the GRE that were statistically equivalent to those achieved with expert human tutors. The research, which involved 2,383 participants, also indicated that in five out of seven GRE subject areas, the top-performing AI tutors surpassed human instructors.

The findings, published on arXiv, stem from a project that developed StudentBench, a platform designed to evaluate AI teaching capabilities and collect data on student-AI interactions. This platform facilitated the collection of over 175,000 messages between students and AI tutors, alongside performance data from participants receiving AI tutoring, human tutoring, or no tutoring. The study focused on quantitative and verbal reasoning sections of the GRE.

Researchers reported that AI tutoring was statistically on par with expert human tutoring for GRE learning gains. Further analysis showed that the most effective AI tutors outperformed human tutors in five of the seven tested GRE domains. This suggests a high level of efficacy for AI in standardized test preparation.

Beyond academic equivalence, the study highlighted a significant cost disparity. One AI tutor achieved comparable learning gains at a cost of approximately $0.0052 per percentage point gained, in stark contrast to the estimated $4.81 per percentage point for human tutoring. This represents a cost difference of roughly 918 times.

The StudentBench project aimed to move beyond simply assessing AI's ability to generate correct answers, focusing instead on actual learning gains. The platform evaluates various aspects of the tutoring experience, including lesson planning, the creation of practice problems, conversational pedagogy, cost-effectiveness, and student engagement.

While the results indicate AI's strong performance in GRE preparation, the authors noted limitations. The study specifically addressed GRE questions and did not establish results for other subjects, age groups, or long-term learning retention. Statistical equivalence at the study level does not guarantee identical support for every student or topic.

The research suggests that AI tutoring presents a scalable and low-cost option for learning, particularly in the context of exam preparation. Future research directions may explore whether these gains transfer to other academic areas beyond standardized tests and how to best integrate AI's capabilities with human oversight and motivation. The StudentBench platform itself allows individuals to experience AI tutoring for free, with data collected for ongoing research.