Ethics and animal numbers: informal analyses, uncertain sample sizes, inefficient replications, and type I errors

Area of Science:

• Biomedical Research
• Animal Ethics
• Statistical Methods

Background:

• Ethical approval for vertebrate animal research requires demonstrating minimal animal use.
• Traditional methods like power analysis and pilot studies have limitations in determining appropriate sample sizes.
• Significance testing often necessitates pre-determined sample sizes, leading to potential overuse or underuse of animals.

Purpose of the Study:

• To explore methods for assuring ethics committees of the minimum necessary animal numbers in research.
• To address the limitations of traditional statistical approaches in sample size determination.
• To introduce sequential stopping rules as a viable alternative for efficient and ethical animal research.

Main Methods:

• Discussion of limitations in power analysis and pilot studies for sample size determination.
• Explanation of issues with traditional significance testing and fixed sample sizes.
• Introduction of sequential stopping rules for incremental data accumulation and analysis.

Main Results:

• Sequential stopping rules allow for interim data analysis without increasing Type I errors (false discoveries).
• This approach ensures statistical significance and replicability.
• It facilitates the assurance of using the minimum number of animals necessary for a scientific goal.

Conclusions:

• Sequential stopping rules offer a robust method for ethical animal research.
• They provide flexibility in data analysis while maintaining statistical integrity.
• This approach supports both scientific rigor and animal welfare in research settings.