T-Test Calculator

📋

Walk-through

How to Use This Calculator

5 steps

1

Choose your test type

Pick One-Sample if you're comparing a single sample's mean against a known or hypothesized value. Pick Two-Sample if you're comparing the means of two independent groups. Pick Paired if you have matched before/after or left/right measurements on the same subjects.

2

Enter your summary statistics

Type in the sample mean(s), standard deviation(s), and sample size(s) for your chosen test — or, for a paired test, the mean and standard deviation of the differences. The calculator updates instantly as you type; no submit button needed.

3

Pick one-tailed or two-tailed

Choose a one-sided alternative before examining the data. For a positive t, the upper-tail p-value is half the two-sided p-value; the lower-tail p-value is one minus that half. Reverse the directions for negative t. A two-sided test checks both directions. Normality or a justified sampling approximation and the stated independence assumptions are still required.

4

Read the t-statistic, df, and p-value

Compare the unrounded p-value with the selected α. Confidence labels follow the selected level and interval type. A small p-value does not establish practical importance; nonrejection does not prove equal means.

5

Calculation workspace

Compare raw data with a supplied summary, preserve named pairs, or compare Welch and pooled tests of the same groups. Central intervals and matching one-sided bounds use the selected α.

⚡

Reference

Formula & Methodology

3 formulas

One-sample and paired t-tests

t = (x̄−µ₀)/(s/√n); paired: t = (d̄−d₀)/(s_d/√n); df = n−1

Compares a sample mean x̄ against a hypothesized population mean µ₀, using the sample standard deviation s and sample size n. A t near 0 means the sample is consistent with µ₀; a large |t| means the sample mean is many standard errors away from µ₀. Treats the n paired differences (after − before, or similar) as a single sample and runs a one-sample t-test against a hypothesized difference of zero, using the mean difference d̄ and the standard deviation of the differences s_d.

Two-sample t-test (Welch's)

t = (x̄₁ − x̄₂) / √(s₁²/n₁ + s₂²/n₂) df = (s₁²/n₁ + s₂²/n₂)² / [(s₁²/n₁)²/(n₁−1) + (s₂²/n₂)²/(n₂−1)]

Welch's t-test compares the means of two independent groups without assuming their variances are equal — the safer default when group sizes or spreads differ. The Welch–Satterthwaite equation gives a (usually non-integer) effective degrees of freedom.

Intervals and directional bounds

t = Δ/SE; CI_central = Δ ± t_(1−α/2,df)SE; L = Δ − t_(1−α,df)SE; U = Δ + t_(1−α,df)SE

Estimates are contrasts: sample mean minus its null, the selected paired mean difference minus its null, or first-group mean minus second-group mean. The opposite paired comparison reverses both contrast and null. Welch allows unequal variances; pooled inference assumes common variance. A one-sided bound has an unbounded opposite endpoint, marked unavailable here. Summary agreement uses the supplied absolute tolerance. Independence and sampling assumptions are not established by these inputs; nonrejection proves neither equality nor practical equivalence.

📖

Glossary

Key Terms Explained

7 terms

t-test A statistical hypothesis test that uses the Student's t-distribution to decide whether an observed difference in means is larger than would be expected from random sampling variation alone.

t-statistic The standardized distance between the observed difference and the hypothesized difference (usually zero), measured in standard errors. Larger |t| means stronger evidence against the null hypothesis.

Degrees of freedom (df) Degrees of freedom describe the model’s remaining information for estimating variability. For a one-sample or paired t-test, df = n − 1. With positive standard error, Welch’s df lies between min(n₁−1, n₂−1) and n₁+n₂−2 and can be fractional. When the standard error is zero, ordinary t inference is unavailable.

p-value The probability of observing a t-statistic at least as extreme as the one calculated, assuming the null hypothesis (no real difference) is true. A small p-value means the observed data would be unusual under the null hypothesis.

Null hypothesis (H₀) The default assumption that there is no real difference — the sample mean equals µ₀, or the two group means (or paired differences) are equal. The t-test either rejects or fails to reject this assumption.

One-tailed vs. two-tailed Choose a one-sided alternative before examining the data. For a positive t, the upper-tail p-value is half the two-sided p-value; the lower-tail p-value is one minus that half. Reverse the directions for negative t. A two-sided test checks both directions. Normality or a justified sampling approximation and the stated independence assumptions are still required.

Statistical significance A result is called statistically significant when its p-value falls below the chosen significance level α (conventionally 0.05) — meaning the observed difference is unlikely to be explained by chance alone.

👥

Scenarios

Real-World Examples

3 worked examples

🏭

Quality control check

One-sample t-test

Sample mean 101.711 g Target mean (µ₀) 100 g Sample std dev 5 g Sample size (n) 25

t = 1.711 with df = 24 gives a two-tailed p ≈ 0.10. Since p is above α = 0.05, this specific sample doesn't provide strong enough evidence that the production line's true average has drifted from the 100 g target — even though the sample average itself is a bit off.

💊

Drug trial (two independent groups)

Two-sample t-test (Welch)

Group 1 mean / SD / n 54.05 / 5 / 13 Group 2 mean / SD / n 50 / 5 / 13

t = 2.064 with df = 24 gives a two-tailed p ≈ 0.05 — right at the conventional significance threshold. For the two groups of 13 observations and equal variances here, Welch’s df is 2(13−1)=24. With positive standard error, df lies between min(n₁−1, n₂−1) and n₁+n₂−2 and may be fractional. Zero standard error makes ordinary t inference unavailable.

🏋️

Before/after training program

Paired t-test

Mean difference (d̄) 2.797 Std dev of differences (s_d) 5 Number of pairs (n) 25

t = 2.797 with df = 24 gives a two-tailed p ≈ 0.01, which is below α = 0.05 — reject the null hypothesis. The average improvement of 2.797 units after training is unlikely to be due to chance.

🔗

Reference

Cite This Calculator

APA & MLA

Use either format to cite this calculator in a paper, report, or resource list.

APA

MLA

📄

Deep Dive

Understanding the T-Test Calculator

3 chapters · 5 min read

A t-test answers one question: is the difference you observed — between a sample and a target, between two groups, or between paired before/after measurements — bigger than random sampling noise would produce? This calculator runs all three common variants from summary statistics you already have (means, standard deviations, sample sizes), so you don't need raw data or statistical software.

Which t-test do I need?

Use the one-sample test when you're comparing a single group's average against a known or hypothesized value (a spec, a historical average, a claimed population mean). Use the two-sample test when you have two separate, independent groups and want to know if their means differ — this calculator uses Welch's version, which does not assume the two groups have equal variance and is the safer default in practice. Use the paired test when the same subjects were measured twice (before/after, left/right, pre/post) — pairing removes subject-to-subject variability so the test is more sensitive to a real effect.

Reading the t-statistic, degrees of freedom, and p-value

The t-statistic measures how many standard errors separate the observed difference from zero (or from µ₀ for the one-sample test) — the larger its absolute value, the less compatible the data are with the null hypothesis of no difference. Degrees of freedom (df) reflect how much independent information went into estimating the variability; for the one-sample and paired tests df = n − 1, while Welch's two-sample test uses the Welch–Satterthwaite formula, which, with positive standard error, gives df between min(n₁−1, n₂−1) and n₁+n₂−2. The value can be fractional. When the standard error is zero, ordinary t inference is unavailable. The p-value converts (t, df) into a probability: it's the chance of seeing a t-statistic this extreme (or more) if the null hypothesis were actually true. Compare it to your significance threshold α (conventionally 0.05) — p < α means reject the null hypothesis, p ≥ α means you don't have enough evidence to do so.

One-tailed vs. two-tailed, and what the assumptions mean

Choose a one-sided alternative before examining the data. For a positive t, the upper-tail p-value is half the two-sided p-value; the lower-tail p-value is one minus that half. Reverse the directions for negative t. A two-sided test checks both directions. Normality or a justified sampling approximation and the stated independence assumptions are still required.

❓

Questions

Frequently Asked Questions

7 questions

Which t-test should I use?+

It depends on your study design, not your preference. One-sample: comparing a single group's mean to a known target. Two-sample: comparing two independent groups. Paired: comparing two measurements taken on the same subjects (before/after, matched pairs). Using the wrong one — for example running an independent two-sample test on paired data — throws away information and gives a less powerful (and sometimes misleading) result.

What's the difference between one-tailed and two-tailed?+

Choose a one-sided alternative before examining the data. For a positive t, the upper-tail p-value is half the two-sided p-value; the lower-tail p-value is one minus that half. Reverse the directions for negative t. A two-sided test checks both directions. Normality or a justified sampling approximation and the stated independence assumptions are still required.

What are degrees of freedom (df), and why isn't Welch's df a whole number?+

Degrees of freedom roughly capture how much independent information you have to estimate variability. For a one-sample or paired test, df = n − 1. Welch's two-sample test instead uses the Welch–Satterthwaite equation, which blends both groups' variances and sample sizes into an effective df that is usually a decimal value — this is expected and is what makes Welch's test valid when the two groups have unequal variances.

What does p < 0.05 actually mean?+

Compare the unrounded p-value with the selected α. Confidence labels follow the selected level and interval type. A small p-value does not establish practical importance; nonrejection does not prove equal means.

Why does the two-sample test use Welch's correction instead of assuming equal variances?+

Student's original pooled two-sample t-test assumes both groups have the same true variance, which is often not the case in practice — especially with unequal sample sizes. Welch's t-test relaxes that assumption and adjusts both the standard error and the degrees of freedom accordingly, making it the more robust, commonly recommended default even when variances happen to be similar.

What assumptions does a t-test make?+

All three variants assume the underlying measurements are approximately normally distributed (or that the sample size is large enough for the sampling distribution of the mean to be approximately normal), and that observations are independent of one another. The paired test additionally assumes the differences themselves are approximately normal. With very small or heavily skewed samples, treat results near the significance threshold with extra caution.

What do the comparison workflows show?+

Contrasts use group/pair order; reversing pairs reverses the null. Welch allows unequal variance; pooling assumes common variance. One-sided bounds leave the other end open. Summary checks use the stated tolerance. Inputs cannot verify sampling assumptions; nonrejection proves neither equality nor equivalence.

Download a one-page PDF of your numbers instantly. Add your email to also get our occasional calculator tips — no spam, unsubscribe anytime. Privacy.

Related calculators

Guides & articles

Glossary terms

Calculator topicT-Test

One-Sample t-Test

Two-Sample t-Test (Welch)

Paired t-Test

How to Use This Calculator

Choose your test type

Enter your summary statistics

Pick one-tailed or two-tailed

Read the t-statistic, df, and p-value

Calculation workspace

Formula & Methodology

Key Terms Explained

Real-World Examples

Quality control check

Drug trial (two independent groups)

Before/after training program

Cite This Calculator

Understanding the T-Test Calculator

Which t-test do I need?

Reading the t-statistic, degrees of freedom, and p-value

One-tailed vs. two-tailed, and what the assumptions mean

Frequently Asked Questions

Get a branded PDF of your results

T-Test Calculator

Test settings

One-Sample t-Test

Two-Sample t-Test (Welch)

Paired t-Test

How to Use This Calculator

Choose your test type

Enter your summary statistics

Pick one-tailed or two-tailed

Read the t-statistic, df, and p-value

Calculation workspace

Formula & Methodology

Key Terms Explained

Real-World Examples

Cite This Calculator

Understanding the T-Test Calculator

Which t-test do I need?

Reading the t-statistic, degrees of freedom, and p-value

One-tailed vs. two-tailed, and what the assumptions mean

Frequently Asked Questions

Get a branded PDF of your results

Keep Exploring

Related calculators

Guides & articles

Glossary terms