Power Analysis in research studies

27 - 29 January 2027 Three days, Belval Campus, Luxembourg Register here

Overview

Designing a sound experiment and choosing the right sample size are among the most important decisions researchers make. Studies with too many subjects waste resources and expose humans or animals to needless procedures and risks. Studies with too few subjects fail to detect the effect under study, so all of the work and risk is wasted. In animal research, a justified sample size is central to the Reduction principle of the 3Rs and is required in ethics applications. Careful planning, sound experimental design and accurate power assessment should be part of every research study plan, ethics application and grant proposal.

This course is aimed at researchers in any field who design or conduct research studies. It starts with how to plan animal experiments using the PREPARE guidelines and with the core principles of experimental design. It then covers power and sample size estimation, with examples for a range of designs and model parameters. It ends with a hands-on exercise: designing an experiment to answer a research question and assessing its power in an RNA-seq setting.

The course will be held on-site over three days, 27–29 January 2027: the afternoon of 27 January, the full day of 28 January and the morning of 29 January.

Learning objectives

  • Plan animal experiments using the PREPARE guidelines
  • Understand the core principles of experimental design (randomisation, controls, experimental unit, replication)
  • Gain a conceptual understanding of statistical power in research projects
  • Know the key factors influencing power
  • Gain experience with tools and software for power analysis
  • Design an experiment to answer a research question and justify its sample size
  • Report power assessments in research study plans, ethics applications and grant proposals

Requirements

Prior Knowledge

No prior requirements are needed, however this course assumes you have some basic knowledge in math and statistics.

Material

You’ll need a laptop to follow the examples and assignments, with the following software installed before the course:

  • G*Power (required): free software for power analysis, available for Windows and macOS. Download the latest version from the G*Power website.
  • R and an IDE (optional): some examples require a recent version of R (4.5.0 or later) and an IDE (RStudio, Positron or VS Code). Please follow the install tutorial to set these up. If you’re not familiar with R, you can skip these examples and the installation.

Preliminary agenda

Day 1 (afternoon): Planning and experimental design

  • Introduction to planning animal experiments with the PREPARE guidelines
  • Introduction to experimental design: experimental units, controls, randomisation and sources of variability

Day 2 (full day)

  • Motivations and principles of power and sample size estimation
  • Main factors influencing power
  • Practical applications using G*Power and pwr R package
  • How to write power and sample size section for grant proposals and peer-reviewed articles
  • Simulations for complex study designs and multiple testing correction methods for omic experiments

Day 3 (morning): From question to design

  • Hands-on exercise: designing an experiment to answer a research question
  • Power assessment in RNA-seq experiments

Registration

Registration is open with limited places. Therefore, it is confirmed on a first-come, first-serve basis.

Address

University of Luxembourg
Belval campus
Maison du Savoir (MSA), room 4.130
Esch-sur-Alzette

Instructor and Contact

Eléonore Schiltz (LCSB, Elixir-Luxembourg)