Planning a research study? This free calculator tells you the minimum number of participants you need to have a good chance of finding a real result — if one exists. Just choose your planned analysis, and the calculator does the rest.
How it works
1
Choose your approach
Tell us whether you have an effect size from previous research, or need to use a general estimate.
2
Pick your analysis
Select the statistical test you are planning — for example, a t-test, ANOVA, or correlation.
3
Get your answer
The calculator instantly shows how many participants you need, with a ready-to-use paragraph for your proposal.
Key terms explained in plain English
Effect size
This is how big the difference, relationship, or change you expect to find. A small effect is subtle and hard to spot. A large effect is obvious. The bigger the effect you expect, the fewer participants you need.
Statistical power
Power is your chance of detecting a real effect if one truly exists. A power of 80% means an 80% chance of finding the effect — this is the traditional minimum. 90% gives a better chance but needs more participants. This calculator shows both.
Significance level (α = .05)
This is the threshold for deciding your result is unlikely to be due to chance. A level of .05 means a 5% risk of a false positive. This calculator uses α = .05 throughout, which is the standard in most research.
Two-tailed test
All calculations use a two-tailed test, meaning the test checks for an effect in either direction. This is the standard approach unless you have a strong reason to test one direction only.
Do you have an effect size from previous research?
The most accurate sample size estimate comes from knowing how large an effect to expect. Choose the option that best describes your situation.
📊Recommended
I have an effect size from previous research
You have found a specific effect size value in a published study, meta-analysis, or pilot study in your research area. You will enter this value directly to get a precise sample size.
You would need a value like:d = 0.43 or r = 0.31 or f = 0.28 or w = 0.35
Where to find this: look in the Results section of a published study similar to yours. The effect size is usually reported alongside the statistical test result.
📐General estimate
I don’t have one — use a general benchmark
You do not have a specific effect size from previous research. This option uses widely accepted benchmarks (Small, Medium, Large) from Cohen (1988) to give you a reasonable starting estimate.
Note: These benchmarks are a useful starting point, but they are general guidelines that do not take into account the specifics of your research area. Where possible, we encourage you to look for effect sizes from previous studies in your field.
Reference: Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum Associates.
What type of analysis are you planning?
Choose the category that best matches your research question.