This guide will help with power analysis for simple experimental designs (e.g., detecting a difference between two or more groups or treatments [t-tests, one-way ANOVA], or the relationship between two variables [correlation, simple regression]). For more complex experimental designs (e.g., mixed designs [time and treatment], many covariates, etc.) I would recommend using the free software G*Power. It is very versatile and easy to use!

There are two potential uses of power analysis.

1. You are planning an experiment (e.g., treatment vs control), and you would like to calculate how many subjects you will need in order to have the best chance of picking up a difference given an effect size of

*x*.2. You have

*n*animals (possibly a small sample), and you want to know what the probability of finding a significant result (given an effect size of*x*)If you are going to calculate your required sample size based on pilot data for the first time, especially for those not familiar with R, I suggest that you first read the section on 'Planning data analysis' on this site. If you are planning to calculate the required sample size based on literature, then read on!