Sample sizes in HFE
By Horacio M Pace-Bedetti, PhD, Principal Human Factors Engineer
Question this answers
What sample size is required for formative and summative human factors studies?
Sample size depends on your research method and objectives. Formative evaluations typically require five to eight participants per user group to detect frequent usability issues. For summative validation studies, the FDA usually looks for at least fifteen participants per user group. Broader methods like surveys or workflow observations require sample sizes matched to the specific research question.
One of the easiest mistakes in Human Factors is treating sample size like a fixed rule instead of a study design decision. There is no single answer, because sample size depends on what you are trying to learn.
For formatives, when the goal is to find as many use errors as possible early, the classic rule is 5 to 8 participants per user group. That is often enough to identify most frequent design and usability issues, especially if you are running several rounds. For summative or validation studies, FDA usually looks for at least 15 per user group. This is based on research from the 90s and early 2000s showing that number is enough to catch most hidden problems, especially in realistic use scenarios.
But here is the thing. These numbers make sense for evaluation methods focused on capturing use errors, but Human factors has a much broader toolbox than that.
The key is to match the sample size to your method and your research question. If you are running in-depth interviews, you may need fewer participants but more sessions. If you are running surveys, you might want a larger sample to spot trends. If you are observing workflow in busy hospital environments, you may need to watch several different shifts or settings.
If your team is pushing back on sample size, do not just quote a number. Bring it back to method and intent. Show what you are trying to learn with the chosen approach. That is the best way to get buy-in and run studies that actually answer the right questions.