Statistical Hypothesis Tester screenshot

Statistical Hypothesis Tester

Run a one-sample t-test, a Welch two-sample t-test or a chi-square goodness-of-fit test, with the statistic, degrees of freedom and p-value.

Privacy First
0 Runs
Updated 2026-11-06
No upload
Instant
No signup

Three tests with a computed p-value

Paste your numbers and choose a test. The one-sample t-test evaluates t = (x̄ − μ₀)/(s/√n) with n − 1 degrees of freedom; the two-sample test is Welch's, which does not assume equal variances; the chi-square test evaluates Σ(O − E)²/E with k − 1 degrees of freedom. The two-sided p-values come from the regularised incomplete beta and gamma functions, computed here rather than looked up in a table, so any sample size or category count is supported.

These are the classical tests under their usual assumptions: independent observations, and for the t-tests an approximately normal sample or a large enough one. The page does not check normality, correct for multiple comparisons, run an ANOVA, or tell you which test your design needs. It reports a p-value, not a decision, and a p-value is not the probability that the null hypothesis is true.

The statistics are computed in this tab from the pasted lists. Nothing is uploaded, no dataset is stored, and the same numbers print the same p-value after a reload.

Three classical tests with a computed p-value: a one-sample t-test t = (x̄ − μ₀)/(s/√n) with n − 1 degrees of freedom, Welch’s two-sample t-test, and a chi-square goodness-of-fit χ² = Σ(O − E)²/E with k − 1 degrees of freedom.

The p-values come from the regularised incomplete beta and gamma functions computed in the page, so any sample size is supported. The page reports a p-value, not a decision, and it does not check the normality or independence the tests assume.

How to use it

Pick the test. Paste the sample for a t-test, or the observed and expected counts for the chi-square, then press Run. The p-value is printed first, then the statistic and its degrees of freedom, then the means or the category count used.

The arithmetic

The one-sample test centres the sample on μ₀ and divides by its standard error. The two-sample test is Welch’s, with its own degrees-of-freedom formula, so unequal variances and unequal sizes are allowed. The chi-square sums the squared gaps between observed and expected, each divided by its expected count.

Where it stops

These are the textbook tests under their usual assumptions: independent observations, and approximately normal samples for the t-tests. The page does not correct for multiple comparisons, run an ANOVA or choose a test for your design. A sample of one, a zero standard deviation, or an expected count of zero is refused because the statistic has no answer.

What to Expect

Best for

  • Checking a t or chi-square test by hand.
  • Getting a p-value for any sample size without a table.

What this tool keeps

  • The two lists are computed in the browser and never posted, so the page works offline and retains nothing.

Sample input

12,15,11,14,13,15,12,14,13,11 against μ₀ = 12

Sample output

p = 0.062903 (two-sided) t = 2.12132, df = 9 mean 13 against μ₀ = 12, n = 10

Frequently asked questions

How do I use it?

Choose the test, paste the sample or the two count lists, then press Run. The p-value, the statistic and its degrees of freedom are printed together.

Is anything uploaded?

No. The arithmetic runs in this tab with JavaScript: no request is made, no account is needed, nothing is stored, and the page keeps working offline once loaded.

What are its limits?

The tests must be used under their assumptions — independent observations and roughly normal samples for the t-tests — and a sample of one or a zero standard deviation is refused.

Developer program

Build custom tools for your store

Join the SDK program to create useful ecommerce utilities for your store.