Experimental Design and Statistics in Animal Research

Experimental design and statistical analysis are tools that are fundamental when performing rigorous scientific investigations. The use of an efficient experimental design, and the associated statistical analysis, is important if the results of the investigation are to be reliable and reproducible, while using the fewest number of animals. This chapter, aimed at scientists without professional statistical support, will discuss some of these tools. Starting with the early planning stages of an experiment the statistical ideas that the researcher should consider are described, for example, formally defining the question they want to answer, sample size calculation, randomisation, blinding, and the use of block designs to reduce variability. Once the experiment has been conducted, a suitable statistical analysis will be required. This chapter will also review some of the more useful tools that any animal researcher should apply in their research. These topics are highlighted and developed with an example of an animal experiment investigating the effect of three therapeutics on irritable bowel disease. This chapter ends with “insider tips” which explore some of the common misconceptions the authors have seen with the application of statistics in animal experiments that should help the reader avoid making similar mistakes in their own work.