Conducting a pilot study
On this page
Why conduct a pilot study?
Animal studies are not usually carried out in isolation, but are part of a larger programme of research. Pilot studies are a useful part of an overall research strategy. A pilot, or feasibility study, is a small experiment designed to test logistics and gather information prior to a larger study. This improves the quality and efficiency of the larger study, and can also bring to light any deficiencies or problems with the design of proposed experiments and allows you to address these before animals, time and resources are used on the larger study. For example, your pilot study may give you vital information on the severity of your planned procedures or treatments.
A pilot study is normally small in comparison with the main experiment, and because of this it can only provide limited information on predicted variability of your outcome measures (both source and size) in your larger study. You are unlikely, for example, to get accurate enough data on variability to use in a power calculation.
To estimate the number of animals needed for a well designed experiment use variability data from a systematic review, or a published study.
Logistical issues which may be revealed by your pilot study
A pilot study can help you resolve logistical issues before you start your main study. It can:
- Check that your instructions to investigators are clear (e.g. how to implement randomisation).
- Check that investigators and technicians have the relevant skills for the planned procedures.
- Check that the equipment works as you expect, and that you know how to use it.
- Check that the experimental animal(s) can perform any tasks required, whether these are physical or cognitive.
- Check the feasibility of your outcome measures (e.g. is it possible to measure the outcome measure in the way you are planning to – including frequency and timing of measurements?).
- Detect a floor or ceiling effect (e.g. if a behavioural task is too difficult or too easy you will get skewed results).
- Assess whether the level of intervention is appropriate (e.g. the dose of a drug).
- Identify adverse effects (pain, suffering, distress or lasting harm) caused by the procedure, and the effectiveness of actions to reduce them (e.g. analgesia dose rate and schedule).
- Define early humane endpoints.
What to do with the data/information from your pilot study
The information you obtain on logistical issues should be incorporated into the design of your main study. The purpose of your pilot study is to assess the feasibility of your experiment, so in most cases you should not present more than summary statistics of your data. In fact, the data from your pilot study may be irrelevant if problems with the methods are discovered.
If your pilot study does not lead to modification of materials or procedures, then it could be possible to include the data in your main study. But be aware that the sampling strategy used to select subjects and the possibility of changes over time must be considered before including your pilot study data. You can include information from your pilot study in publications, even if the main study differs, as this can inform design of future experiments. If the pilot study leads to large changes in design, a second pilot study may be necessary, or in some cases the main study may have to be abandoned.
References and further reading
- Lancaster GA et al. (2004). Design and analysis of pilot studies: recommendations for good practice. J Eval Clin Pract 10(2): 307-12. doi: 10.1111/j..2002.384.doc.x.
- Ruxton GD and Colegrave N (2006). Experimental Design for the Life Sciences. 2nd edition. Oxford University Press.
- Cochran WG and Cox GM (1992). Experimental Designs. 2nd edition. John Wiley & Sons.
- Altman DG (1991). Practical Statistics for Medical Research. 1st edition. Chapman & Hall.
- Festing MF et al. (2016). The design of animal experiments: reducing the use of animals in research through better experimental design. 2nd edition. Sage Publishing.