Survey Weighting Adjustment Calculator
Compute the post-stratification weight for a survey subgroup that is over- or under-represented relative to the population.
How it works
weight = target population share ÷ sample share · sample share = group count ÷ total
Post-stratification weighting corrects a sample that does not match the population on a known characteristic such as age, gender or region. Each respondent in a group is multiplied by the ratio of the group's true population share to its share in your sample, so an under-represented group is scaled up and an over-represented one is scaled down. Applied across all cells, this brings weighted estimates back in line with the population.
Worked example
If 450 of 1,000 respondents are women (45% of the sample) but the population is 52% women, the weight is 0.52 ÷ 0.45 = 1.156. Each female respondent then counts as 1.156 people in weighted totals.
Frequently asked questions
Do large weights cause problems?
Yes. Weights far from 1 inflate the variance of your estimates and reduce the effective sample size. Weights above about 3 or below 0.3 are usually trimmed or capped to keep results stable.
Can I weight on several variables at once?
Yes — this is often done with raking (iterative proportional fitting), which adjusts weights repeatedly until the sample matches the population on every margin simultaneously. This calculator handles one cell at a time.
How does weighting affect the margin of error?
Weighting increases the design effect, so a weighted survey has a larger effective margin of error than an unweighted one of the same size. Report the effective sample size when quoting precision.
Related calculators
This calculator is for educational purposes. Double-check important results before acting on them.