Course Retire statistical significance: a world beyond p<0.05

The course C312/ Retire statistical significance: a world beyond p<0.05 is being offered by the Graduate School of Health, Aarhus University, 2026.

Criteria for participation: University degree in medicine, dentistry, nursing, or Master’s degree in other fields and/or postgraduate research fellows (PhD students and research-year medical students).

Requirements for participation: University degree in medicine, dentistry, nursing, or master’s degree in other fields and/or postgraduate research fellows (PhD students and research-year medical students). Basic biostatistical knowledge is assumed.

Aim: To avoid misuse and misinterpretations of statistical significance testing

Learning outcomes: 

• To define point estimates, confidence intervals, and p-values.
• To identify mistaken inferences due to misinterpretations of significance tests.
• To recognize the clinical context's importance in interpreting effect estimates.
• To apply Apps to aid in the interpretation of effect estimates.
• To evaluate treatment effects considering both systematic and random errors.
• To reflect on the approach to data interpretation in your research environments.

Workload: The full workload of the course is expected to be 20 hours (incl. 5 hours pre-course reading)

Content: 

• Review of statistical and epidemiological concepts
  • Statistical inference
  • Internal validity: systematic vs. random errors
  • Hypothesis testing, p-value, point estimates, and confidence intervals
  • Interpretation of effect estimates
• The clinical context
  • Clinical statistical significance
  • The Null result
  • Clinical trial considerations
  • Sample size calculations
• Misinterpretations of significance testing
  • The epidemiologist’s perspective
  • The statistician’s perspective
  • Key papers, persons, and arguments in the debate
• Advanced topics
  • Type 1 vs. type 2 errors: which is worst?
  • 99% CI vs. 95% CI: which is best?
  • Confidence Limit Ratios
  • P-value functions
  • Multiple comparisons: A place for p-values?
  • Is Bayesian inference the solution?
• Expert panel debate (online): Pro et cons*
• When Peer Reviewers Push for P-values:
  • o The role of editors and reviewers
  • o How to reply, resist, and refuse
• Study activities:
  • o The Conclusion Generator: Apply your own results
  • o Multiple choice assessment: Alignment with learning objectives

The content is relevant for all health researchers, but the examples provided during the course will relate to clinical and epidemiological research and not basic (experimental) research.

* Expert panelists

2025 (Fall): Blakeley B McShane (Northwestern University); Erik Parner (Aarhus University); Prof. Stephen Senn (University of Dundee)
2024 (Fall): Ronald L. Wasserstein (Executive director, the American Statistical Association); Gerd Gigerenzer (Vice-president European Research Council); Erik Parner (Aarhus University).
2024 (Spring): Stuart Pocock (LSHTM); John Ioannidis (Stanford University); Erik Parner (Aarhus University).
2023 (Fall): Mathew Fox (Boston University); Valen Johnson (Texas A&M University); Erik Parner (Aarhus University).
2023 (Spring): Valentin Amrhein (University of Basel); John Ioannidis (Stanford University); Erik Parner (Aarhus University).
2022: Kenneth Rothman (Boston University); Timothy Lash (Emory University); Andreas Stang (University Hospital Essen).

Instructors: Professor Morten Schmidt, MD, PhD, DMSc and Professor Erik Parner, MSc, PhD.

Venue: Aarhus University, Aarhus.

Participation in the course is without cost for:

• PhD students, Health Research Year students from Aarhus University
• PhD students enrolled at partner universities of the Nordoc collaboration
• PhD students from other institutions in the open market agreement for PhD courses

Course dates:

• 02 December 2026 08:00 - 15:30
• 03 December 2026 08:00 - 15:30