Date
Publisher
arXiv
In this paper, we investigate whether current state-of-the-art large language models (LLMs) are effective as AI tutors and whether they demonstrate pedagogical abilities necessary for good AI tutoring in educational dialogues. Previous efforts towards evaluation have been limited to subjective protocols and benchmarks. To bridge this gap, we propose a unified evaluation taxonomy with eight pedagogical dimensions based on key learning sciences principles, which is designed to assess the pedagogical value of LLM-powered AI tutor responses grounded in student mistakes or confusion in the mathematical domain. We release MRBench -- a new evaluation benchmark containing 192 conversations and 1,596 responses from seven state-of-the-art LLM-based and human tutors, providing gold annotations for eight pedagogical dimensions. We assess reliability of the popular Prometheus2 LLM as an evaluator and analyze each tutor's pedagogical abilities, highlighting which LLMs are good tutors and which ones are more suitable as question-answering systems. We believe that the presented taxonomy, benchmark, and human-annotated labels will streamline the evaluation process and help track the progress in AI tutors' development.
Study at a glance
- Question: How effectively do AI models demonstrate teaching skills when helping students work through math mistakes and confusion?
- Setting and data: Math tutoring conversations: 192 dialogues and 1,596 responses from seven AI and human tutors.
- Method: Evaluates and compares AI and human tutor responses across eight teaching dimensions using human annotations and an AI evaluator.
Why it matters
It remains unclear whether large language models demonstrate the pedagogical abilities necessary for effective tutoring in educational dialogues, as previous evaluation efforts have been limited to subjective protocols and benchmarks.
Key findings
- In mathematics dialogues, some leading AI models function as effective tutors, whereas others are only suitable as question-answering systems.
- AI tutor performance can be evaluated across eight teaching dimensions grounded in learning sciences principles when addressing student confusion.
- Seven leading AI and human tutors demonstrated distinct teaching abilities across 192 conversations and 1,596 responses to student errors.
Summary drafted with AI from the paper’s abstract and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
