Best Practices for Behavior Study Design
This Page includes general practices that should be followed for optimal behavior study design and data collection conditions. These practices should be applied, where appropriate, to all behavioral studies.
- Variability is the death of statistical analysis – and variability can be greater in behavioral data due to many external factors that influence behavioral performance. These include things like environmental noises, issues in the colony room, time of day, experimenter, etc. Therefore, consistency is key and reducing the number of batches of mice run per study helps to mitigate variability across batches. Variability can also be mitigated with increased sample size.
- Consider the effect size of your experimental manipulation to determine the appropriate sample size for your study. Group sizes of 20 or less are only powered to see large effects. To observe moderate effects, group samples of around 50 are necessary. To observe anything with a small effect, 100 samples or more per group will be required in your experiment. The Core provides consulting on power analyses to aid in determining needed study sample size.
- Number of mice in the cage can influence performance, especially in social behavior assays. Avoid single housing mice whenever possible. Three mice to a cage is ideal for male mice for their dominance hierarchical structure.
- Avoid rehousing males over P35.