How can we turn clinical questions into meaningful statistical evidence?

Understanding and applying statistics is essential for clinical research and evidence-based medicine. The Clinical Course Biostatistics combines a concise refresher on descriptive and inferential statistics with practical data analysis in R. Students develop the skills needed for their master’s thesis and future clinical research, from formulating a research question and selecting appropriate methods to interpreting findings and assessing their clinical relevance.

After the course, students will be able to:

1. Understand and critically apply statistical principles in a medical context.
2. Select and apply suitable statistical methods to clinical questions.
3. Use R efficiently to analyse medical data.
4. Conduct an empirical study, from the research question to the presentation of results.
5. Interpret statistical findings correctly and assess their clinical relevance.
6. Present scientific results clearly and discuss them in relation to existing research.

Course content

Topics include measurement scales, descriptive statistics, correlation, t-tests, analysis of variance, linear and logistic regression, and chi-square tests. Particular attention is given to checking statistical assumptions and interpreting p-values, effect sizes, and confidence intervals. Students also explore how to structure scientific findings around the introduction, methods, results, and discussion of a research article.

Course structure

The course comprises four sessions of two teaching units each. At its core is a research project conducted in small groups: students formulate a clinical or biomedical question, select suitable statistical methods, analyse their data in R, and interpret the findings in their medical context.

The project culminates in a short presentation and discussion of each group’s empirical analysis or study, grounded in the relevant scientific literature.

Tools and practical resources

• Statistical analysis: R, available via CRAN, and RStudio.
• AI companions for statistical discussion: ChatGPT, Qwen3.6-27B, and Posit Assistant with Ollama. These can support dialogue about statistical concepts and assumptions, with explanations checked against the textbook.
• Experimental workflow presented by the lecturer: R, Visual Studio Code, Claude Code, and Quarto, including Claude Code integration with VS Code.
• Alternatives for simple statistical calculations: Microsoft Excel and Google Sheets.