Survey statistics workbench
Four browser-only calculators for common survey analyses. Enter counts or summary statistics; your inputs stay on this device.
NPS score and comparison, a Wilson confidence interval for one proportion, Pearson’s chi-square test of independence, or Welch’s independent-samples t-test.
NPS calculator
Use 0–6 as detractors, 7–8 as passives and 9–10 as promoters. NPS = % promoters − % detractors.
Compare two NPS scores
Approximate independent-group comparison using a normal approximation and unadjusted two-sided test.
Confidence interval for a proportion
Wilson score interval for one survey proportion, generally better behaved than a basic normal interval near 0% or 100%.
Chi-square test of independence
Enter a 2 × 2 table of observed counts. The calculator reports expected counts and warns when the large-sample approximation is weak.
Welch’s t-test for two means
Compare independent groups without assuming equal variances. Enter each group’s mean, standard deviation and sample size.
Methods and limits
NPS comparison uses an approximate standard error for the difference between independent NPS estimates. Confidence intervals for proportions use the Wilson score method. Chi-square uses Pearson’s statistic and the chi-square distribution. The t-test uses Welch–Satterthwaite degrees of freedom and Student’s t distribution. These methods assume independent observations; they do not correct for complex survey design, weighting, clustering, repeated measures or multiple testing. Statistical significance is not practical importance.
Need wording examples? Browse the survey question library. Questions and calculations stay in your browser; no AI service is involved.
References
- Wilson, E. B. (1927), “Probable inference, the law of succession, and statistical inference.” JASA 22(158), 209–212.
- Pearson, K. (1900), “On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling.”
- Welch, B. L. (1947), “The generalization of ‘Student’s’ problem when several different population variances are involved.” Biometrika 34(1–2), 28–35.
Frequently asked questions
Can this compare paired NPS waves?
No. The comparison shown is an approximate independent-group test. Repeated measurements on the same respondents need a paired analysis.
Does a non-significant result prove there is no difference?
No. Review the confidence interval and the size of effects that would matter for the decision; a result above the chosen p-value threshold does not establish that groups are identical.
Do these methods handle weighting or clustering?
No. These methods assume independent simple random observations. Complex survey designs need design-aware standard errors and analysis.
Run research end to end. Keep the knowledge working.
One AI-native operating system for market research and insight professionals — from study design and evidence generation to agents, institutional knowledge, delivery and action.