Scholarlink

<p>This study explores the integration of ChatGPT as a learning tool in a second-year mathematical statistics course, comparing its effectiveness to traditional recorded lectures. Two challenging topics, order statistics and moment generating functions, were chosen to assess student learning outcomes. A mixed-methods approach was employed, involving quiz performance and survey feedback. Quantitative analysis revealed no significant difference in quiz scores between the ChatGPT group and the traditional learning group (Mann–Whitney test: <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(p\)</EquationSource> <EquationSource Format="MATHML"><math> <mi>p</mi> </math></EquationSource> </InlineEquation>-value = 0.98), with average scores of 35.48% and 58.39% for Quiz 1 and Quiz 2 respectively in the ChatGPT group, compared to 34.86% and 58.25% in the traditional group. Qualitative feedback highlighted high satisfaction with ChatGPT’s explanations (78.5% satisfaction with instructions and 81.3% overall experience rating). However, concerns about reliability (68.8%) and academic integrity (18.8%) were noted. Traditional lectures were valued for clarity (85.1% satisfaction) and overall experience (88.8% rated positively). This research underscores ChatGPT's potential to complement traditional teaching methods, offering flexibility and personalized support while addressing challenges in AI-assisted education. The findings inform future strategies for integrating AI into higher education, enhancing accessibility, engagement, and learning outcomes.</p>

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

AI Meets Academia: Enhancing Statistics Education with ChatGPT in Undergraduate Courses

  • 期刊论文

  • 发表日期：

    2025年7月1日

查看全文
  • Luai Al-Labadi,
  • Omidali Aghababaei Jazi,
  • Masoud Ataei,
  • Kairui Bao,
  • Ruodie Yu

摘要

This study explores the integration of ChatGPT as a learning tool in a second-year mathematical statistics course, comparing its effectiveness to traditional recorded lectures. Two challenging topics, order statistics and moment generating functions, were chosen to assess student learning outcomes. A mixed-methods approach was employed, involving quiz performance and survey feedback. Quantitative analysis revealed no significant difference in quiz scores between the ChatGPT group and the traditional learning group (Mann–Whitney test: \(p\) p -value = 0.98), with average scores of 35.48% and 58.39% for Quiz 1 and Quiz 2 respectively in the ChatGPT group, compared to 34.86% and 58.25% in the traditional group. Qualitative feedback highlighted high satisfaction with ChatGPT’s explanations (78.5% satisfaction with instructions and 81.3% overall experience rating). However, concerns about reliability (68.8%) and academic integrity (18.8%) were noted. Traditional lectures were valued for clarity (85.1% satisfaction) and overall experience (88.8% rated positively). This research underscores ChatGPT's potential to complement traditional teaching methods, offering flexibility and personalized support while addressing challenges in AI-assisted education. The findings inform future strategies for integrating AI into higher education, enhancing accessibility, engagement, and learning outcomes.