With Artificial Intelligence (AI) being a novel topic in education, AI tools have become increasingly widespread among students, with massive potential in aiding student learning. It offers ways to improve learning in various educational domains, including mathematics. ChatGPT is a Large Language Model (LLM) with the potential to foster learning. LLMs excel in logical reasoning, problem-solving, and abstraction, allowing them to be capable of teaching mathematics. LLMs are also interactive, personalizable, and flexible. This study investigates the use of ChatGPT and Prompt Engineering (PE) to improve student performance on the topic of conic sections in the Senior High School PreCalculus course. Applying PE could produce more accurate and insightful responses when fed to LLMs. The AI-assisted learning tool is composed of engineered, easy-to-use, and click-to-copy prompts presented through a website, which serves as input to ChatGPT. The tool is tested quantitatively and qualitatively. Statistical tests on pre-test and post-test results, user experience analyses, and chat log analyses are used to determine the viability of this tool. Results revealed that students’ performance through test scores increased significantly, meaningfully, and fairly. Participants also reported the overall effectiveness of the tool as a learning aid. Chat logs show that mistakes were observed only rarely, and exclusively in responses that tackle lengthy and complex concepts. The results of this study show that the AI-assisted learning tool improved students’ performance on the topic of conic sections in the PreCalculus course.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Exploring the Application of AI-Assisted Learning in Improving Performance in Mathematics

  • Nathanael Theodore M. Fenomeno,
  • Adrienne R. Cabochan,
  • Chloe Jewel S. Rivera,
  • Adrian Matthew L. Alcantara,
  • Shirley B. Chu

摘要

With Artificial Intelligence (AI) being a novel topic in education, AI tools have become increasingly widespread among students, with massive potential in aiding student learning. It offers ways to improve learning in various educational domains, including mathematics. ChatGPT is a Large Language Model (LLM) with the potential to foster learning. LLMs excel in logical reasoning, problem-solving, and abstraction, allowing them to be capable of teaching mathematics. LLMs are also interactive, personalizable, and flexible. This study investigates the use of ChatGPT and Prompt Engineering (PE) to improve student performance on the topic of conic sections in the Senior High School PreCalculus course. Applying PE could produce more accurate and insightful responses when fed to LLMs. The AI-assisted learning tool is composed of engineered, easy-to-use, and click-to-copy prompts presented through a website, which serves as input to ChatGPT. The tool is tested quantitatively and qualitatively. Statistical tests on pre-test and post-test results, user experience analyses, and chat log analyses are used to determine the viability of this tool. Results revealed that students’ performance through test scores increased significantly, meaningfully, and fairly. Participants also reported the overall effectiveness of the tool as a learning aid. Chat logs show that mistakes were observed only rarely, and exclusively in responses that tackle lengthy and complex concepts. The results of this study show that the AI-assisted learning tool improved students’ performance on the topic of conic sections in the PreCalculus course.