Solving the Self-regulated Learning Problem: Exploring the Performance of ChatGPT in Mathematics

Pin Hui Li, Hsin Yu Lee, Yu Ping Cheng, Andreja Istenič Starčič, Yueh Min Huang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

22 Citations (Scopus)

Abstract

In flipped math classrooms, chatbots are commonly used to assist students and provide personalized learning to improve self-regulation issues that students face when learning through online resources at home. ChatGPT, the state-of-the-art natural language model, has been tested in this study to explore its ability to impact middle school students’ math learning since middle school math is a crucial stage that affects future math learning success. The study tested ChatGPT’s accuracy by using it to answer questions from Taiwan’s past education examinations, and the accuracy rate was found to be as high as 90% (A+). Moreover, compared to most studies that developed chatbots for a single unit or course, this study found that ChatGPT’s accuracy in each of the six major areas of mathematics education in Taiwan exceeded 80% (A). The results indicate that ChatGPT is an excellent learning tool that can improve students’ self-regulation issues and has the potential to impact middle school math education.

Original languageEnglish
Title of host publicationInnovative Technologies and Learning - 6th International Conference, ICITL 2023, Proceedings
EditorsYueh-Min Huang, Tânia Rocha
PublisherSpringer Science and Business Media Deutschland GmbH
Pages77-86
Number of pages10
ISBN (Print)9783031401121
DOIs
Publication statusPublished - 2023
EventProceedings of the 6th International Conference on Innovative Technologies and Learning, ICITL 2023 - Porto, Portugal
Duration: 2023 Aug 282023 Aug 30

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14099 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

ConferenceProceedings of the 6th International Conference on Innovative Technologies and Learning, ICITL 2023
Country/TerritoryPortugal
CityPorto
Period23-08-2823-08-30

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Solving the Self-regulated Learning Problem: Exploring the Performance of ChatGPT in Mathematics'. Together they form a unique fingerprint.

Cite this