About t-Test Calculator
Free t-test calculators and plain guides for students, researchers and analysts, with results checked against R and SciPy.
Checked on November 21, 2026. Results computed by the site's engine and tested against R and SciPy.
What this site does
It runs t-tests and related tests correctly and shows its working: the exact p-value, a confidence interval, a named effect size, an APA sentence and code that reproduces the result in R or Python. The guides explain which test to use and how to report it.
Who is behind it
t-Test Calculator is run from New Zealand. James Graham is the editor, and the site is published by Crucible.
James Graham has spent 17 years in analytics, quantitative modeling and commercial strategy, including quantitative modeling roles at Lotto NZ and Synlait, and 15 years building search-led products that make practical knowledge easier to reach. He holds a BA from the University of Auckland in Human Geography, with spatial analysis and mapping, and Biological Anthropology. Crucible is his company.
Neither James Graham nor Crucible claims formal statistics qualifications, and no outside statistician has checked the pages. Instead, every calculation is tested against published reference results, listed in full on the methodology page, so you can check the numbers yourself.
How content is made, including AI
Pages are drafted with the help of AI tools and then reviewed and edited by James Graham. AI is not used as a source of numbers: every figure on every page is computed by the site's statistics engine when the site is built, so prose cannot contain a number of its own. Statements about software behavior and methods cite their published sources, such as R's source code, SciPy's documentation, NIST and peer-reviewed papers.
How the calculators are tested
Before anything is published, automated checks must pass: 30 reference cases from NIST, R's reference output, SciPy and textbooks, more than 8,000 stored SciPy values, and a check of every page. See the methodology for each case and our value.
Editorial standards
- Interpretation uses neutral wording: "evidence of a difference at the 5% level", rather than a claim of proof, and "no evidence of a difference" rather than "the means are equal".
- Effect sizes are always named, and benchmarks are called rough conventions.
- Where software or textbooks disagree, we show both and say which we follow.
- When we have not confirmed something ourselves, the page says Not confirmed.
- No business pays to appear, and there are no affiliate links.
Publisher
The site is published by Crucible Investments Limited, a New Zealand limited company, NZBN 9429053412676, company number 9401778. Write to [email protected]. Crucible's website is crucible.co.nz.
How the site is funded
The site is funded by advertising. Ads appear below results and within guides, never inside a calculator.
Add the calculator to your site
You are welcome to embed the t-test calculator. It carries no ads, stores nothing and shows a small credit link. Copy this code:
<iframe src="https://ttestcalculator.com/embed/" title="T-test calculator" width="100%" height="1500" style="border:0;max-width:560px" loading="lazy"></iframe>