19  Pilot

Although not essential, it is often a good idea to conduct a formal pilot of the intervention. Pilot results can inform feasibility and identify modifications needed in the design of the larger study. Note that the results of a pilot cannot inform about the effectiveness of the intervention per se due to small sample sizes. However, it can serve several purposes:

• It is a final check to see if the program as designed is feasible and has a reasonable chance of success.
• It may help understand in greater detail the pathways by which the program might have an impact, and refine the intermediate and final outcome measures.
• It may help to test the data collection instrument.

If you would like to know more, read The role and interpretation of pilot studies in impact evaluation research by Shagun Sabarwa. The case study underlines the importance of having broad pilot studies in informing the causal hypothesis, study design, key variables of interest measurement and finally the decision of whether or not to proceed with a randomised evaluation.

For more information, see Glennerster and Takavarasha (2013).

Glennerster, R., and Takavarasha, K. (2013). The experimental approach. In Running randomized evaluations: A practical guide (pp. 1–23). Princeton University Press. https://www.jstor.org/stable/j.ctt4cgd52.5