Sample-size calculator
Survey sample-size calculator for MEAL teams
Work out how many people to survey for an assessment or evaluation — from population, confidence level, margin of error, design effect and cluster size — using the standard proportion-based formula.
Every step is shown, so you (or your donor's reviewer) can see exactly how the number was reached.
Calculate your sample size
Use 50% if you're not sure — it gives the largest, safest sample.
Leave blank for a very large or unknown population.
1.0 = simple random sampling. Use about 1.5 for two-stage household cluster surveys.
Completed responses needed
385 respondents
- Base sample (simple random)
- 384.1
How this was calculated
How the calculator works
It uses Cochran's formula for a proportion — n₀ = z² · p(1−p) / e² — then applies your design effect (a multiplier for cluster sampling), the finite population correction when you enter a population, and finally inflates for expected non-response. Sample sizes are always rounded up, so you never under-power the survey.
Choosing your inputs
Confidence level
95% is the standard for most surveys; 90% is sometimes accepted for rapid assessments.
Margin of error
±5 percentage points is typical. Halving it roughly quadruples the sample.
Expected proportion
The indicator's expected value. If you don't know it, 50% is the safest choice.
Design effect
Use 1.0 for simple random sampling; about 1.5 for two-stage household cluster surveys (DHS/SMART).
Response rate
Inflates the number you need to approach so you still hit the target after refusals and absences.
Not sure what to measure? Choose your indicators first, then size the sample.
Browse the Indicator LibraryGood to know
This calculator uses the standard proportion-based formula for a single estimate — the right approach for most coverage, KAP and multi-sector needs assessments. For stratified, multi-stage or comparison-of-groups designs, confirm the sample with your M&E advisor.
Sample-size calculator
From sample design to clean data.
Humanitarian Ops runs assessments, MEAL and beneficiary data end to end — so the sample you plan here flows straight into collection and reporting. See how it fits your team — we usually reply within 24 hours.