MTMasterTool

T-Test Calculator

Calculate t-test using your data.

Loading calculator…

What this tool does

Calculate t-test using your data. Independent samples, unequal-variance Welch test; assess distribution, outliers and sampling design.

How to use the T-Test Calculator

  1. Enter or select sample a: comma or whitespace-separated numbers.
  2. Enter or select sample b: comma or whitespace-separated numbers.
  3. Read the calculated result; change any measurement to compare alternatives.

Formula

Welch t=(mean A−mean B)/sqrt(sA²/nA+sB²/nB); df=(a+b)²/(a²/(nA−1)+b²/(nB−1)); two-sided p=regularized beta(df/(df+t²),df/2,1/2)
a
Sample A: comma or whitespace-separated numbers
b
Sample B: comma or whitespace-separated numbers

Independent samples, unequal-variance Welch test; assess distribution, outliers and sampling design.

Worked example

For t-test calculator, the following measurements illustrate the exact method: Sample A: comma or whitespace-separated numbers: 1,2,3,4,5; Sample B: comma or whitespace-separated numbers: 2,4,6,8,10.

Inputs

  • Sample A: comma or whitespace-separated numbers1,2,3,4,5
  • Sample B: comma or whitespace-separated numbers2,4,6,8,10

Result

  • Welch t statistic-1.9
  • Degrees of freedom5.88
  • Two-sided p-value0.11

Results explained

Welch t statistic
Welch t statistic from the formula above. Independent samples, unequal-variance Welch test; assess distribution, outliers and sampling design.
Degrees of freedom
Degrees of freedom from the formula above. Independent samples, unequal-variance Welch test; assess distribution, outliers and sampling design.
Two-sided p-value
Two-sided p-value from the formula above. Independent samples, unequal-variance Welch test; assess distribution, outliers and sampling design.

Frequently asked questions

Welch t=(mean A−mean B)/sqrt(sA²/nA+sB²/nB); df=(a+b)²/(a²/(nA−1)+b²/(nB−1)); two-sided p=regularized beta(df/(df+t²),df/2,1/2)

Independent samples, unequal-variance Welch test; assess distribution, outliers and sampling design.

Enter numbers only, separated as described in the labels. Check sample units and remove missing values rather than substituting zero.

No. Statistical inference depends on sampling design, independence, model fit and interpretation; a p-value is not the probability that a hypothesis is true.

No. Calculations and numerical distribution algorithms run entirely in the browser.