In this blog post, Dr. Wei Lin Teoh, Associate Professor at the School of Mathematical and Computer Sciences (Malaysia) discusses how real-world applications can be used to transform statistical learning.
Traditionally, statistical learning at the university level has often been taught through theory-heavy lectures, focusing on mathematical derivations, and standardised exercises. While this builds a strong conceptual foundation, it can sometimes feel disconnected from real-world applications. Students may master techniques but struggle to see how these methods solve practical problems.
My teaching philosophy takes a different approach. I emphasise student engagement and learning through research and discovery. I aim to create an inclusive and practical learning environment that sparks curiosity and equips students to apply statistical thinking to real problems. Instead of treating statistics as abstract, I integrate real datasets and inquiry-based activities to make learning meaningful. I believe learning is most effective when students engage in a supportive environment that fosters collaboration, deeper understanding, and confidence.
Reimagining Statistical Methods: What Did I Do and Why?
Undergraduate statistical education faces increasing pressure to stay relevant in a world where data drives decisions across every sector. Statistics is not just about formulae, it is the language we use to interpret data, quantify uncertainty, and turn raw information into meaningful insights. When students only encounter idealised examples, they miss the complexity and unpredictability of real data, which is where true statistical thinking is tested.
The challenge is that traditional models were designed for an era when data was scarce and controlled. Today, data is abundant, messy, and often unstructured. Employers and researchers expect graduates to handle large datasets, use modern analytical tools, and communicate findings clearly to diverse audiences. This shift demands a curriculum that moves beyond theory to practical application.
In 2023, I played a central role in redesigning F79PS Further Statistical Methods, a course for third- and fourth-year students across Heriot-Watt’s campuses in the UK, Malaysia, and Dubai. The redesign focused on making learning more engaging and relevant by incorporating authentic datasets, computational tools, and projects based on real data, preparing students for the complexities of modern data analysis.
The course was redesigned based on two key principles:
- Authentic learning
- Active pedagogies
We moved away from traditional lectures and introduced a flipped classroom approach (Howell, 2021; Rincón, Munárriz & Ruiz, 2025). Students learned the basic concepts before class and used class time for active, hands-on problem-solving. This approach enabled deep engagement without compromising course coverage.
We also changed the course materials to focus on real-life case studies and simulated industrial scenarios. Students used SPSS and R software (widely used tools for statistical analysis and data modelling) to explore real datasets, helping them connect theory with practice. We updated the assessments to reflect workplace scenarios, helping students develop practical skills that are directly applicable to their future careers.
Real-World Integration: Bringing Statistics to Life
One of the main goals of the new course design was to show how statistics are used in the real world. Students worked on real datasets and were given problems similar to those faced by professionals.
For example, students analysed data from fast-food restaurants to uncover patterns in nutritional content. They selected key variables, such as calories, fat, and cholesterol, and applied K-means clustering, a technique that groups similar items based on shared characteristics. After interpreting the resulting clusters, they transformed these insights into practical recommendations aimed at improving menu options.
Another task asked students to study medical data on children with insulin-dependent diabetes. They applied Locally Estimated Scatterplot Smoothing (LOESS), a technique that fits smooth curves to data by focusing on local patterns rather than assuming a single global trend. This allowed them to explore how age and acidity levels influenced C-peptide levels in the blood.
These weekly tasks gave students the chance to learn by doing. They also encouraged teamwork and discussion, creating a classroom environment where students helped and learned from each other.
Enhancing Student Outcomes
The changes yielded visible improvements in both student experience and academic outcomes. In 2024, student satisfaction increased, and many performed better in their assessments. In their feedback, students said that the real-world examples made the course more useful and enjoyable. They described the learning experience as “inclusive”, “practical”, and “engaging.”
Students also reported greater confidence in using statistical tools and tackling unfamiliar problems. Some projects even yielded insights with real-world relevance. For example, analysing global blood pressure data from 199 countries (1992–2020) to identify hypertension patterns, and predicting abalone age from physical measurements as an alternative to counting shell rings. While primarily for learning, these projects showed how classroom work can spark ideas that extend beyond the university.
A highlight was when two students won second place in the 2024 ASEAN Data Science Explorers National Finals. Their project addressed Ischaemic Heart Disease, one of Malaysia’s leading causes of death. They proposed data-driven solutions to promote heart health, including tools to encourage healthier eating habits across ASEAN.
Broader Lessons: How Can Others Benefit?
The transformation of F79PS Further Statistical Methods offers several strategies applicable for other educators:
- Connect theory with practice by using real-world data and realistic tasks.
- Use active learning methods, like flipped classrooms and group work, to make learning more engaging.
- Encourage students to learn together, creating a supportive and collaborative environment.
- Let students explore and discover, using research-based projects to spark interest and build confidence.
By applying these ideas, educators can create dynamic, student-centred learning environments that are both academically strong and practical.
Final Thoughts
Teaching statistics today is about more than just explaining formulae. It is about helping students learn how to think critically, solve real problems, and work together. By combining real-world tasks and learning through discovery, we can give students the tools that they need to succeed not only in exams, but also in life.
References:
Howell, R.A. (2021). Engaging students in education for sustainable development: The benefits of active learning, reflective practices and flipped classroom pedagogies. Journal of Cleaner Production, 325, article 129318. https://doi.org/10.1016/j.jclepro.2021.129318
Rincón Y.R., Munárriz, A., & Ruiz, A.M. (2025). Flipped classroom or flip to foster self-regulation competencies in mathematics in economics and business students. International Journal of Educational Research, 130, article 102556. https://doi.org/10.1016/j.ijer.2025.102556
Image credits:
Photo by Wei Lin Teoh
Social media:
LinkedIn: https://www.linkedin.com/in/dr-teoh-wei-lin-phd-a650159a/
