ToolsConductScience tool
Welch's t-testFree in-browser calculator

General Stats Calculator.

Welch's t-test with Cohen's d and Pearson correlation with Fisher z CI. Enter raw data or summary statistics. Data never leaves your browser.

PrivateData stays in your browser
LiveNo sign-up required
Open methodFormula shown on this page
CitableMethods and citation included

Calculator

Results update in place

Try it out

Load example General Stats data to see the full workflow

When to use

  • Compare means between two independent groups using Welch's t-test
  • Assess linear association between two continuous variables with Pearson correlation
  • Report effect sizes (Cohen's d or R-squared) with confidence intervals
  • Generate APA-style methods paragraphs for basic statistical analyses
  • Analyze data from published summary statistics (n, mean, SD) when raw data is unavailable

Do not use for

  • Paired or repeated-measures designs — a paired t-test or repeated-measures ANOVA is needed
  • More than two groups — use ANOVA (not available in this tool)
  • Non-normal data with small samples — consider a non-parametric test (Mann-Whitney U)
  • Method comparison / agreement analysis — use the Method Comparison Analyzer

Always use Welch's t-test as the default

Welch's t-test does not assume equal variances and performs well even when variances are equal. Student's t-test is only slightly more powerful under strict equal-variance conditions, which are rarely verifiable in practice.

Report effect size, not just p-value

A statistically significant p-value tells you the effect is unlikely to be zero, but not whether the effect is meaningful. Always report Cohen's d (for t-tests) or R-squared (for correlation) with confidence intervals.

Fisher z transformation is essential for correlation CIs

The sampling distribution of Pearson r is skewed, especially when |r| is large. Computing a CI directly on r gives inaccurate coverage. The Fisher z transformation normalizes the distribution for proper CI computation.

Correlation does not imply causation

Pearson r measures linear association only. A strong correlation can be driven by confounders, restricted range, or outliers. Always inspect a scatter plot and consider the study design before interpreting r as evidence of a relationship.

1

Method

Welch's t-test uses the Welch-Satterthwaite degrees of freedom approximation, which does not assume equal variances. Its p-value and the 95% CI for the mean difference come from the exact t distribution at those degrees of freedom. Cohen's d is computed as the mean difference divided by pooled SD, with a 95% CI of d ± 1.96 times its approximate standard error. Pearson correlation uses the Fisher z transformation (z = 0.5 ln((1+r)/(1-r))) for CI computation, then back-transforms to the r scale. Its p-value comes from the exact t distribution with n − 2 degrees of freedom. All calculations run client-side in the browser.

2

Published

Published 2026-11-10. Calculations support planning and documentation; verify procurement decisions against manufacturer specifications or institutional SOPs.

3

How to cite

How to Cite

ConductScience General Stats Calculator. ConductScience, Inc. 2026. Available at: https://conductscience.com/tools/general-stats-calculator

Welch BL. The generalization of "Student's" problem when several different population variances are involved. Biometrika. 1947;34(1-2):28-35. doi:10.1093/biomet/34.1-2.28

Statistical Testing Fundamentals

Statistical hypothesis testing follows a standard workflow:

1. State hypotheses: H₀ (no difference/association) vs H₁ (difference/association exists) 2. Choose test: Based on data type, design, and assumptions 3. Compute test statistic: t, F, χ², or r 4. Report p-value and effect size: The p-value alone is insufficient — always report the effect size and confidence interval

Modern reporting standards (APA, CONSORT) require: test statistic, degrees of freedom, p-value, effect size, and CI. This tool generates all of these automatically.

Frequently asked

What is Welch's t-test?
Welch's t-test compares the means of two independent groups without assuming equal variances. Unlike Student's t-test, it adjusts the degrees of freedom using the Welch-Satterthwaite equation, making it robust to unequal sample sizes and variances. Most statisticians recommend always using Welch's test as the default, since it performs well even when variances are equal.
What is Cohen's d?
Cohen's d is a standardized measure of the difference between two group means, expressed in pooled standard deviation units: d = (M₁ − M₂) / SD_pooled. Cohen's benchmarks: d = 0.2 (small), d = 0.5 (medium), d = 0.8 (large). Always report d with a confidence interval — the point estimate alone is insufficient.
What is Pearson correlation?
Pearson correlation coefficient (r) measures the linear association between two continuous variables. Values range from −1 (perfect negative) to +1 (perfect positive). R2\text{R}^{2} gives the proportion of variance in Y explained by X. Correlation does not imply causation, and r is sensitive to outliers and restricted range.
Why use Fisher z for correlation CI?
The sampling distribution of r is skewed (especially when |r| is large), so computing a CI directly on r gives inaccurate coverage. The Fisher z transformation z = ½ ln((1+r)/(1−r)) has an approximately normal distribution, enabling accurate CI computation. The CI is then back-transformed to the r scale.
Should I use raw data or summary statistics?
Use raw data whenever possible — it allows the tool to verify assumptions and compute exact statistics. Summary statistics (n, mean, SD) are useful when you only have published results from another study, but you lose the ability to check for outliers or non-normality.
325
Free tools
1,200+
Institutions
100%
Client-side
0
Uploads required