AI-Driven Assessment of Human Tutors: Linking Training Performance to Real-Life Practice

20d ago · Global · primary source: export.arxiv.org

A new AI-driven system can evaluate human tutors by analyzing transcripts of their real-life teaching sessions, linking performance during training directly to classroom outcomes, according to research published on arXiv [1]. The system, detailed in a paper submitted June 17, uses the generative AI model Gemini-2.5-pro to assess both open-ended responses during training and authentic tutoring sessions, measuring how well tutors transfer skills to real-world application [1]. The study involved 86 human tutors who provided remote math instruction and completed six scenario-based lessons, resulting in an average learning gain of 7.4% [1]. Researchers used mixed-effects models across 405 session-to-lesson pairs and found that training performance significantly predicted real-life transcript scores, with an effect size of 0.25 SD [1]. A comparison of models indicated that averaging open-response and multiple-choice performance during training best predicted real-life tutoring quality, though open responses alone were more predictive than multiple-choice scores [1]. Exploratory analysis showed that after training, tutors encountered pedagogical opportunities to apply their skills more frequently, rising from 61.1% to 68.9%, and their execution quality within those opportunities improved from 65.5% to 68.1% [1]. An interrupted time series analysis suggested these gains were part of a gradual trend rather than an immediate result of the training intervention [1]. The authors released open datasets, AI prompts, and scoring rubrics to support transparency and reproducibility [1]. The work appears on arXiv, a preprint platform that has integrated with Hugging Face Spaces to allow researchers to link interactive demos directly alongside papers [6][7]. The broader field of educational technology draws on disciplines including artificial intelligence and computer science to facilitate learning, with AI applications spanning language translation, decision-making, and e-commerce [3][4].

commentarymodel-releaseresearch-paper

Background sources we checked (10)
  • arxiv.org ↗ There exist numerous tutor training platforms. However, few provide AI-driven training and evaluation for human tutors based on real-life performance. We present an AI-driven system that assesses both open responses during training and authentic real-life tutoring. Unlike platfor…
  • en.wikipedia.org ↗ Artificial intelligence is the capability of computational systems to perform tasks that are typically associated with human intelligence, such as learning, reasoning, problem-solving, perception, and decision-making. Artificial intelligence has been used in applications througho…
  • en.wikipedia.org ↗ Educational technology (often abbreviated as edtech) encompasses computer hardware, software, along with educational theories and practices, used to facilitate learning and teaching. When referred to by its abbreviation, "EdTech," it often denotes the industry of companies that d…
  • en.wikipedia.org ↗ The artificial intelligence (AI) market in India is projected to reach $8 billion by 2025, growing at 40% CAGR from 2020 to 2025. This growth is part of the broader AI boom, a global period of rapid technological advancements with India being pioneer starting in the early 2010s w…
  • huggingface.co ↗ Hugging Face Machine Learning Demos on arXiv Back to Articles ... # Hugging Face Machine Learning Demos on arXiv Published November 17, 2022 Update on GitHub Upvote 1 - - - - - Abubakar Abid abidlabs Follow …
  • info.arxiv.org ↗ ## Hugging Face Spaces ... Hugging Face code repositories, About Hugging Face ... Collaborators: Abubakar Abid, Omar Sanseviero, Ahsen Khaliq, and the Hugging Face team ... Hugging Face Spaces includes links to demos created by the community or the authors themselves. By going to…
  • huggingface.co ↗ Demos on Hugging Face Spaces allow a wide audience to try out state-of-the-art machine learning research without writing any code. Hugging Face and ArXiv have collaborated to embed these demos directly along side papers on ArXiv! ... Thanks to this integration, users can now find…
  • en.wikipedia.org ↗ Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence (AI) company that develops large language models (LLMs). Based in Hangzhou, Zhejiang, DeepSeek is owned and funded by High-Flyer, a Chin…
  • en.wikipedia.org ↗ A large language model (LLM) is a type of machine learning model designed for natural language processing tasks such as language generation. LLMs are language models with many parameters, and are trained with self-supervised learning on a vast amount of text.…
  • en.wikipedia.org ↗ Douwe Kiela is a Dutch-American research scientist and entrepreneur working in the field of artificial intelligence with a focus on machine learning and natural language processing. He is a research scientist director at Google DeepMind. He previously co-founded and served as CEO…

Sources

Spot something wrong? Report an issue