Stratified Sample Allocation Calculator
Split a total sample across strata by proportional or equal allocation and get the interviews needed in each stratum.
How it works
Proportional: nₕ = n · Nₕ ÷ N · Equal: nₕ = n ÷ k
Stratified sampling divides the population into non-overlapping groups (strata) and samples within each, guaranteeing every group is represented. Proportional allocation gives each stratum a share of the sample matching its share of the population, which needs no weighting afterward. Equal allocation puts the same number in each stratum — useful when you need reliable estimates for small strata, at the cost of needing weights to recombine them.
Worked example
Strata of 5,000, 3,000 and 2,000 (N = 10,000) with a total sample of 500, proportional: 500·0.5 = 250, 500·0.3 = 150, 500·0.2 = 100. Equal allocation would instead give about 167 in each.
Sources: Stratified sampling — Wikipedia
Frequently asked questions
When should I use equal instead of proportional allocation?
Use equal allocation when you need to report on each stratum separately with similar precision — for example comparing four regions. A small stratum would otherwise get too few interviews under proportional allocation.
What about Neyman (optimal) allocation?
Neyman allocation also weights each stratum by its variability, sending more of the sample to strata with higher variance. It is the most efficient method but requires an estimate of each stratum's standard deviation, which this calculator does not use.
Do I need to weight the results afterward?
Proportional allocation is self-weighting, so simple tallies are already representative. Equal or Neyman allocation is not, so you must apply stratum weights when combining the strata into an overall estimate.
Related calculators
This calculator is for educational purposes. Double-check important results before acting on them.