Scholarlink

Background <p>Medical schools are under increasing pressure to teach larger cohorts while maintaining an excellent student experience. The University of Manchester replaced Problem-Based Learning (PBL) with Team-Based Learning (TBL) in 2023 to meet growing demand for medical education. While the educational effectiveness of TBL is well established, this study evaluated how the routine data generated within large-scale TBL could be potentially used to support early intervention, monitor curriculum quality, and validate peer assessment processes.</p> Methods <p>We conducted a prospective mixed-methods evaluation of two consecutive Year-1 medical cohorts at a large UK medical school (2023–24, <i>n</i> = 435; 2024–25, <i>n</i> = 426). Data included TBL individual readiness assurance test (iRAT) scores, student experience surveys, operational measures, and end-of-semester examination results. In the second year, the 50 lowest-performing students after four teaching themes were identified using iRAT data and offered tailored academic and wellbeing support through their academic advisors.</p> Results <p>TBL maintained high student satisfaction (74% rated sessions “excellent”) whilst simultaneously reducing weekly staff facilitator hours by 63% (160&#xa0;h vs. 59&#xa0;h) and room use by 84% (80&#xa0;h of bookings vs. 13). Student individual weekly performance, as measured by iRAT scores, was strongly predictive of summative performance (R²≈0.38–0.41) and this finding is reproducible over consecutive years. Theme-level variation in iRAT performance also identified differences in content difficulty and alignment, providing a programme-level indicator of curriculum quality and areas of weaker delivery. Early identification and intervention of struggling students increased semester performance by 20% (<i>p</i> &lt; 0.0001) among the bottom 50 students, improved summative exam marks by 5%, and reduced exam failure rate by 41%, without inflating whole cohort attainment.</p> Conclusions <p>This study shows how routine TBL data can function as a scalable analytics system to support early intervention, continuous curriculum monitoring, and quality assurance of peer assessment, while also reducing operational demands. These findings provide a transferable model for institutions seeking to expand capacity while maintaining educational quality and equity.</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

Teaching at scale: a mixed-methods study of team-based learning analytics to enable early support in medical education

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月15日

查看全文
  • Elizabeth Sheader,
  • Lisa Donlan,
  • Connor Allen,
  • Nathan Betteridge,
  • Michael Dixon,
  • Ingrid Gouldsborough,
  • Ruth Grady,
  • Margaret Kingston,
  • Valerie Kouskoff,
  • Kerry Parris,
  • Tristan Pocock,
  • Maiedha Raza,
  • Paul Shore,
  • Nicholas Stafford,
  • Cathy Tournier,
  • Michelle Webb,
  • Michael P. Smith

摘要

Background

Medical schools are under increasing pressure to teach larger cohorts while maintaining an excellent student experience. The University of Manchester replaced Problem-Based Learning (PBL) with Team-Based Learning (TBL) in 2023 to meet growing demand for medical education. While the educational effectiveness of TBL is well established, this study evaluated how the routine data generated within large-scale TBL could be potentially used to support early intervention, monitor curriculum quality, and validate peer assessment processes.

Methods

We conducted a prospective mixed-methods evaluation of two consecutive Year-1 medical cohorts at a large UK medical school (2023–24, n = 435; 2024–25, n = 426). Data included TBL individual readiness assurance test (iRAT) scores, student experience surveys, operational measures, and end-of-semester examination results. In the second year, the 50 lowest-performing students after four teaching themes were identified using iRAT data and offered tailored academic and wellbeing support through their academic advisors.

Results

TBL maintained high student satisfaction (74% rated sessions “excellent”) whilst simultaneously reducing weekly staff facilitator hours by 63% (160 h vs. 59 h) and room use by 84% (80 h of bookings vs. 13). Student individual weekly performance, as measured by iRAT scores, was strongly predictive of summative performance (R²≈0.38–0.41) and this finding is reproducible over consecutive years. Theme-level variation in iRAT performance also identified differences in content difficulty and alignment, providing a programme-level indicator of curriculum quality and areas of weaker delivery. Early identification and intervention of struggling students increased semester performance by 20% (p < 0.0001) among the bottom 50 students, improved summative exam marks by 5%, and reduced exam failure rate by 41%, without inflating whole cohort attainment.

Conclusions

This study shows how routine TBL data can function as a scalable analytics system to support early intervention, continuous curriculum monitoring, and quality assurance of peer assessment, while also reducing operational demands. These findings provide a transferable model for institutions seeking to expand capacity while maintaining educational quality and equity.