
I've got the power - Main findings from the Lund Biomedicine ReproducibiliTea Journal Club
Jun 26
2 min read
1
21
0

Why are power and sample size estimations so hard and scary?
In the latest Lund Biomedicine ReproducibiliTea Journal Club, Magnus Pettersson, senior biostatistican at Statistikkonsulterna Väst, led a discussion about power and sample size estimations.
Science is supposed to benefit mankind and shall be done in an ethical manner. Wasting resources and harming test subjects (whether humans or animals) is never acceptable.
Key takeaways from the meeting
It is complicated not knowing what the results will be in advance, thus forcing one to rely on assumptions regarding variation, results, and compliance. But, some creativity and smart questions can guide you to the answers you need. Other studies can give a rough estimate to start with.
Statistics is mathematics, and we expect a single and correct solution. But in power analysis, we are in a grey zone, where we obtain a range of reasonable solutions to consider. Therefore, we need to weigh information on accepted risks versus a health economic cost-benefit analysis.
A limited budget—whether with regard to time or sample size—is always a reality. However, a well-made power analysis can be a good argument for justifying the budget.
Using the opportunity to develop a good power analysis will not only give you a head start to an efficient study—it will also encourage a thorough discussion about potential pitfalls, structure the data collection, and plan the analysis ahead.
Join the Conversation
Our upcoming journal club meetings, set to resume in Fall 2025, will continue exploring critical research questions. If you are a researcher, student, or science enthusiast, join us to engage in meaningful dialogue about improving scientific methods and research integrity.
Register your interest to receive updates and sign up for upcoming ReproducibiliTea Journal Club meetings.
Speaker and guest author

Magnus Pettersson, fil.lic, is senior biostatistician and CEO of Statistikkonsulterna Väst AB. He has a strong interest in education, ethics, and quality in healthcare and the life sciences.
Editor

Sean Kim, PhD, is a medical writer at AdvanSci Research Solutions. He has been a medical writer since 2006, writing extensively on therapeutics for a wide range of areas.