Identify threshold concepts in introductory statistics and effectiveness evaluation of using Shiny Apps

TILT program: Teaching and Learning Development Grant (TLDG)

Principal Investigator: Wei (Becky) Lin, lecturer, Department of Statistics and Actuarial Science, Faculty of Science

Project team: Tim Swartz, consultant, professor, Department of Statistics and Actuarial Science, Faculty of Science; Ivy Lendero, research assistant, School of Criminology, Faculty of Arts and Social Sciences; Mariana Toniolo Barrios, TILT research assistant

Timeframe: April 2022 - May 2023

TILT Support: $1600 and up to 110 hours of TILT research assistant time

Course addressed: STAT 201 and STAT 203 -- Introductory Statistics

Final Report: View Becky Wei Lin's final report (PDF)

Description:

This project examined how identifying threshold concepts in introductory statistics and using the interactive web-based application Shiny Apps can address persistent teaching and learning challenges. 

The project introduced the Fun Apps of Statistical Tools (FAST) initiative, which uses Shiny Apps to support learning without requiring coding skills. The project had two phases. Phase I focused on identifying threshold concepts, defined as core ideas that are conceptually transformative but often troublesome for learners. Threshold concepts were identified through consultations with senior faculty members, analysis of midterm and final exam results, student interviews, and teaching assistant feedback. This process revealed common areas of difficulty, including hypothesis testing, confidence intervals, p-values, probability distributions, and interpreting statistical relationships.

Based on these findings, the interactive Shiny Apps was developed for key topics in descriptive statistics, probability, inference for means and proportions, regression, chi-square tests, and ANOVA. Phase II evaluated the effectiveness of these tools using exam performance comparisons, student surveys, and course evaluations. Results showed significant improvement in student performance after the introduction of Shiny Apps, with higher mean and median exam scores (p < 0.01). Survey responses indicated overwhelmingly positive student perceptions, with most students reporting that the app was easy to use, reduced anxiety, and helped them focus on understanding concepts rather than performing tedious calculations.

Overall, the project demonstrates that identifying threshold concepts and addressing them through interactive, visual tools can significantly improve learning outcomes and student experience in introductory statistics. Shiny Apps provide an accessible, engaging, and pedagogically effective approach for supporting conceptual understanding at scale.

Questions addressed:

• What are the threshold concepts in introductory statistics courses STAT 201 and STAT 203?
• Which statistical concepts do students struggle with most based on examination performance?
• How effective are Shiny Apps in improving student performance in introductory statistics?
• What are students' perceptions of using interactive Shiny Apps for learning statistics?
• How can interactive web applications help students learn statistics without encountering coding barriers?

Knowledge sharing: Project findings have been shared with colleagues including access to the Shiny Apps. Two faculty members have incorporated Shiny Apps into their course design and teaching. Becky has also been invited by the Statistical Education Section of Statistical Society of Canada (SSC) to present project findings (2023, Ottaway). There are also plans to give a formal presentation to the department and to write paper.

Keywords: Threshold concepts, introductory statistics, interactive web applications, use of technology, Shiny Apps, effectiveness, student engagement, R programming, statistical education, learning barriers