Currently, the scientific literature is critical of the translatability of in vivo research results and of the reproducibility of research conducted with animals. There are certainly solutions on what should be done to improve the translatability and reproducibility of animal experiments.
Among the solutions, statistics and experimental design play a central role to reach better translatability and reproducibility, in addition to contribute to the Reduction arm of the ThreeRs. This course addresses the central question of how to design your experiments and which test to perform given your specific research questions.
This course offers lectures, exercises and discussions on these topics.
STATISTICS FOR ANIMAL EXPERIMENTS
Reserch design:
- Confouding factors
- Randomization
- Blinding
- Defining outcomes and endpoints
Hypothesis tests:
- How to select your statistical test ?
- One and two-sided tests
- Pitfalls and caveats in hypothesis testing
- Multiple testing and p-hacking
- Student t-test
- ANOVA
THE EXPERIMENTAL DESIGN: HOW TO GENERALIZE THE RESULTS
How to calculate the «right» sample size:
- Power calculation
- Calculating power using simulations
- Post-hoc power calculation
Experimental design illustrated by practical examples:
- Randomized, blinded, controlled experiments
- Formal experimental design: Completely randomized, Randomized complete block, Factorial design, Examples, advantages and limitations
Duration:
3 mornings, 4 - 11 - 25 October 2024 CEST time zone