Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Statistics›Bartlett's Test for Homogeneity of Variances
Hypothesis testVariance homogeneity

Bartlett's Test for Homogeneity of Variances

Also known as: Bartlett's Chi-Square Test, Test for Equality of Variances, Bartlett's Homogeneity Test, Varyans Homojenliği Testi

Bartlett's Test is a classical parametric procedure for testing whether two or more independent groups share a common population variance. Introduced by Maurice Stevenson Bartlett in 1937, it formalises the null hypothesis that all group variances are equal by constructing a chi-square statistic from the ratio of pooled to individual group variances. It is a standard pre-analysis step before applying ANOVA or other procedures whose validity depends on the homoscedasticity assumption.

ScholarGate
  1. Hypothesis test
  2. v1
  3. 1 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Bartlett's Test
Fligner-Killeen TestOne-way ANOVALevene and Brown-Forsyth…

When to use it

Bartlett's Test is appropriate when comparing variances across two or more independent, normally distributed groups before conducting ANOVA or similar analyses. It assumes that each group's observations are independently and normally distributed; the test is notably sensitive to departures from normality, which can inflate Type I error. When normality is uncertain, Levene's or Fligner-Killeen tests are recommended alternatives. Best suited to continuous data with reasonably large within-group sample sizes.

Strengths & limitations

Strengths
  • Optimal power under normality: when the normality assumption holds, Bartlett's test is asymptotically the most powerful test for variance equality.
  • Chi-square reference distribution: the analytic chi-square approximation avoids resampling, making the test fast and easy to implement.
  • Handles multiple groups: extends naturally to k groups simultaneously, unlike simple two-sample F-tests.
  • Widely supported: available in virtually all major statistical software packages and well-established in the literature.
Limitations
  • Sensitive to non-normality: even mild departures from normality can substantially inflate the Type I error rate, making the test unreliable without prior normality verification.
  • Poor performance with small samples: the chi-square approximation deteriorates when group sample sizes are small, leading to inaccurate p-values.
  • Does not identify which groups differ: a significant result indicates at least one group variance differs, but post-hoc procedures are needed to localise the difference.
  • Restricted to continuous data: the test is not appropriate for ordinal or categorical variables.

Frequently asked

How does Bartlett's test differ from Levene's test?

Bartlett's test is based on likelihood-ratio principles and assumes normality, giving it optimal power under that assumption but making it sensitive to non-normal data. Levene's test uses absolute deviations from group means or medians and is far more robust to non-normality, making it the preferred choice when the distribution of residuals is uncertain.

What should I do if Bartlett's test is significant?

A significant result means the equal-variance assumption is likely violated. Depending on context, you may apply variance-stabilising transformations, switch to Welch's ANOVA (which does not require homoscedasticity), or use non-parametric alternatives. Post-hoc inspection of group standard deviations helps identify which groups drive the heterogeneity.

Can Bartlett's test be used with only two groups?

Yes. With two groups, Bartlett's test is equivalent to the two-sample F-test for equality of variances. However, even for two groups the test retains its sensitivity to non-normality, so Levene's or Fligner-Killeen tests remain better choices whenever the normality of each group cannot be confirmed.

Sources

  1. Bartlett, M. S. (1937). Properties of sufficiency and statistical tests. Proceedings of the Royal Society of London. Series A, 160(901), 268–282. DOI: 10.1098/rspa.1937.0109 ↗

How to cite this page

ScholarGate. (2026, June 2). Bartlett's Test for Homogeneity of Variances. ScholarGate. https://scholargate.app/en/statistics/bartlett-test

Related methods

Fligner-Killeen TestOne-way ANOVA

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Fligner-Killeen TestStatistics↔ compare
  • One-way ANOVAStatistics↔ compare
Compare side by side →

Referenced by

Fligner-Killeen TestLevene and Brown-Forsythe Test

Similar methods

Levene and Brown-Forsythe TestFligner-Killeen TestOne-way ANOVAWelch ANOVAAnalysis of Variance (ANOVA)Robust ANOVAWelch t-testIndependent t-test

Related reference concepts

Multivariate Analysis of VarianceData Distribution and NormalityStatistical Hypothesis TestingLikelihood-Ratio TestsChi-Squared and Fisher Exact TestsCentral Limit Theorem

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Bartlett's Test (Bartlett's Test for Homogeneity of Variances). Retrieved 2026-07-21 from https://scholargate.app/en/statistics/bartlett-test · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Maurice Stevenson Bartlett
Year
1937
Type
Parametric variance homogeneity test
Subfamily
Variance homogeneity
Null Hypothesis
All group population variances are equal
Test Statistic
Chi-square approximation
Related methods
Fligner-Killeen TestOne-way ANOVA
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account