Why is experimental design important?

Good experimental design is important in all research, it helps to ensure the data collection, data analysis and conclusions from a study, are valid (true). Good experimental design is essential in research and fundamental to scientific progress. Research projects are written up into ‘papers’ which are submitted for publication in peer-reviewed scientific journals. These papers are cited and used to inform future research should be as accurate as possible, ensuring the studies are robustly designed, conducted and reported is therefore essential. 

Sadly, there are examples of poorly designed research published in even the most respected scientific journals, despite systems of review designed to avoid this. Where research does not have a robust design, it can lead to bias in the results or poor data/statistical analyses leading to incorrect conclusions. These research outcomes are misleading and can render the research of little value.  

At FRAME we advocate for good experimental design and believe that good science across all experiments, animal and non-animal, is vital to helping reduce and ultimately replace animal research. Here we explain why. 

Why is experimental design training important in animal research? 

Experimental design is key to all scientific research, and all those conducting research should have a solid understanding of the basics of planning a study and how to avoid bias. This starts in primary school and should continue through postgraduate study. There is room for more to be done through education and training to ensure future researchers have the knowledge and skills to correctly plan any research study, including animal studies. Whilst animal research is still happening, and we are working to reduce reliance on animals in research, supporting good experimental design is important for several reasons:  

• It ensures that any animal research studies are planned and carried out correctly. This includes correctly implementing the 3Rs to replace and reduce animal use and potential suffering as much as possible.
• Ensures animal’s lives are not used in poorly planned research which may then be repeated (using yet more animals).
• It builds up a reliable evidence base of areas of research where animal studies do not translate to humans. (For example, in areas of human disease research)
• It ensures studies using modern, non-animal approaches are planned and conducted correctly to help build the evidence base for new approaches to research questions, increase confidence in them and lead to greater uptake and advancement where animals may still be used. 

In regulated preclinical animal research, and indeed clinical research using patients, understanding experimental design is important for scientific reasons and helps avoid, for example, the common practice of focusing on one gender in a study to reduce variability in results due to sex, which is frequently highlighted as a poor scientific choice missing out vital differences between how different genders respond to drugs and disease. As well as scientific concerns, poor experimental design raises concerns around wasted money, time and resources when valuable funding is essentially wasted on the conducting and publication of poorly planned studies. The peer-review process should reduce the chance of this research being published, but this does not always work – see this article ‘Poorly designed animal experiments in the spotlight’.

There are also ethical issues when animals are being used in poorly designed studies which therefore have questionable value. Animal researchers must correctly implement the 3Rs in their studies, one of which is Reduction. This requires researchers to plan and use the minimum number of animals necessary to answer their research question validly. There is a risk here of going too far the other way and using too few animals, which can make it impossible to draw reliable conclusions and produce outcomes which are neither reproducible or useful. Ultimately this means no benefit accrues from the project and it will likely be repeated in the future, using yet more animals. In order to ensure the minimum number of animals necessary are used, without compromising the outcomes of their study, researchers can apply power analysis to their experimental design. Power analysis is a statistical equation that allows a minimum sample size to be detected, whilst ensuring effect size can be reliably detected. Read more about power analysis.