Stats Made Simple: How to Learn Statistics

How to Learn Statistics (Without Getting Overwhelmed): A Starter Guide for Young Rheumatologists and Rheumatology Researchers

Statistics are the cornerstone of modern evidence-based medicine. Moving from clinical observations to robust evidence requires the application of appropriate statistical methods. Already at the design stage of a study, statistics play a crucial role: ensuring adequate power, choosing the right tests, and avoiding bias. Importantly, it is not only about p-values; statistics also determine the real-life strength of associations, for example, through hazard ratios or odds ratios. 

Although medical or other university curricula may include statistics, many young rheumatologists or researchers find themselves overwhelmed when confronted with the complexity of statistical methods in the literature. Yet statistical literacy is essential — both for clinical practice, to judge the quality of diagnostic and therapeutic studies, and for basic and translational research, where advanced methods are increasingly required for data analysis. This article provides guidance on how to get started.

Start with the basics

Most universities and hospitals offer introductory statistics courses, which remain the best starting point. Ask colleagues or supervisors which ones are most suitable and when to take them. To apply statistics effectively, it is also essential to learn at least one statistical software program well; for example, R, SPSS, Stata, or Python. As an addition to live courses, several reliable online platforms provide excellent introductions from leading universities. Most of these courses can be followed free of charge, with a subscription needed only if you want a certificate. 

How to get unstuck

Once you have understood the basic statistical concepts and have the basic skills in statistical software, the real challenge begins: applying methods to your own data. Do not wait until you have mastered all the courses you have attended. You learn statistics by performing tasks and discovering which tools are most relevant for your work. Do not try to master everything at once, but focus on what you need. Pair up with a colleague to stay motivated or join a statistical group at your university.

A common barrier at the start is not the statistics themselves, but data handling: converting variables into the right format, restructuring datasets, or cleaning data. Colleagues can often help, but increasingly, artificial intelligence tools can support your learning. They can explain difficult concepts in simple language, generate examples, or create practice questions. They are also useful for writing or debugging R or Python code. The key is in the prompting: be specific in what you ask. (See Textbox 1 for starter prompts.) Remember, however, that AI makes mistakes and is not a substitute for proper training or critical thinking.

Out-of-the box learning

Not everyone learns statistics the same way, and traditional methods can feel dry. Here are some alternative strategies:

• Podcasts provide accessible introductions during commutes or downtime.
• Gamification through learning apps encourages active practice via short daily challenges.
• Micro-learning through reviewing one concept per day from a recent article promotes consistency over cramming.
• Exploratory practice with open datasets (e.g. WHO data, Kaggle medical datasets) allows hands-on experimentation with regressions and visualisations.
• Journal clubs that critically examine the statistical methods used in published studies are particularly effective, especially when participants explain or teach the concepts to peers.

Conclusion

Do not be discouraged: learning statistics is a process, not a hurdle to get over. With the right resources and a bit of creativity, it can even become fun. And to support you, the EMEUNET newsletter will continue this series to provide practical articles on statistical methods in rheumatology in the coming months.