The Importance of Basic Statistics in Early Career Research
For veterinary professionals beginning a career in research, statistics can sometimes feel like an area best left to statisticians. However, a an understanding of basic statistics is an essential part of not only conducting research but interpreting and communicating the findings to colleagues and owners. Whether investigating treatment outcomes, disease prevalence, diagnostic tests, animal welfare, or clinical risk factors, statistical thinking helps researchers turn observations into meaningful evidence.
One of the most important roles of statistics is in study design. Good research begins long before data are analysed. Early-career veterinary researchers need to consider questions such as how many animals should be included, how animals will be selected, what variables should be measured, and whether comparison or control groups are required. Understanding concepts such as sampling, randomisation, bias, variability, and statistical power can help researchers design studies that are both scientifically robust and ethically responsible. This is particularly important when clinical decisions will often be based on or inferred from this research.
Basic statistics are equally important when describing data. Measures such as the mean, median, range, standard deviation, percentages, and confidence intervals allow you to summarise your findings clearly. However, selecting the appropriate measure matters. For example, a small number of animals with exceptionally long hospital stays could substantially influence the mean, making the median a more informative description of a typical patient. Understanding the distribution and variability of data therefore helps prevent misleading conclusions. Understanding when to use median over mean is critically important and where to couple standard deviation or interquartile ranges, these are common errors I see in my editorial roles.
Statistical knowledge also supports the appropriate use and interpretation of hypothesis testing. You will commonly come across p-values, but statistical significance should not be confused with clinical importance. A statistically significant difference between two veterinary treatments may be too small to influence clinical decision-making. Conversely, a clinically meaningful difference may fail to reach statistical significance in a small study. Looking beyond the p-value to effect sizes, confidence intervals, sample size, and study design provides a much more complete interpretation of the evidence.
Finally, basic statistical literacy makes veterinary professionals better users of research. Clinicians routinely encounter published studies that may influence decisions about diagnosis, treatment, prevention and ultimately, animal welfare. Being able to recognise inadequate sample sizes, inappropriate analyses, confounding factors, or overinterpretation of results allows veterinarians to assess evidence critically rather than accepting conclusions at face value.
You do not need to become a statistician. You do, however, need enough statistical understanding to ask the right questions, recognise limitations, collaborate effectively with statistical experts, and communicate results responsibly. Developing these skills early provides a strong foundation for high-quality veterinary research and, ultimately, better evidence-based care for animals.
I would highly recommend using R-Studio for any statistical work you may want to do and whilst at first it is daunting, AI can be a massive help but only if used responsibly and with some understanding of what you are aiming to achieve and how you should go about getting there! There are some excellent resources out there to help with R and one of my favourites is:
Pius, E.H. and R. R for Health Data Science. Available at: https://argoshare.is.ed.ac.uk/healthyr_book/
Some other useful resources are: