Skip to content
Omnikalc

Bartlett test calculator

Enter the size and the standard deviation of three groups to get the Bartlett statistic, the p-value and the pooled standard deviation.

Updated on
Your inputs

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Results

Bartlett statistic

0.998

p-value (2 degrees of freedom)
0.6071
No significant difference between the variances
Pooled standard deviation
4.499
Ratio of the largest to the smallest standard deviation
1.34

The Bartlett test calculator gives the Bartlett statistic, the p-value and the pooled standard deviation for three groups, to check whether their variances can be treated as equal.

How the calculation works

  • Enter the size and the standard deviation of each of the three groups.
  • The calculator pools the variances and builds the corrected statistic.
  • It gives the statistic, the p-value on 2 degrees of freedom, the pooled standard deviation and the ratio of the extreme standard deviations.

Example

Groups of 12, 15 and 10 with standard deviations of 4.2, 5.1 and 3.8 have a pooled variance of 4.499². The Bartlett statistic is 0.998 with 2 degrees of freedom, so the p-value is 0.6071. The ratio of the largest to the smallest standard deviation is 1.34.

Why it matters

The calculator helps students, researchers and quality engineers to check the equal variance assumption before an ANOVA or a pooled t-test, using only the summary statistics of the groups.

Good to know

A common rule of thumb says that the assumption is fine when the largest standard deviation is less than twice the smallest, and the calculator shows this ratio. The Bartlett test is the most powerful test of equal variances when the data are truly normal, but it is fragile otherwise, so look at a histogram or a normal plot first. With very large samples, even a small and harmless difference becomes significant, so combine the p-value with the ratio and with the effect on your conclusions. For two groups only, the F test of two variances is the usual choice. The calculator handles three groups, and the formula extends to more groups by adding terms.

The Welch ANOVA calculator compares means when variances differ, the ANOVA calculator assumes equal variances, the chi-square distribution calculator gives the p-values and the control chart limits calculator monitors a process.

Frequently asked questions

What does the Bartlett test check?

It tests the null hypothesis that all groups have the same variance. The statistic compares the log of the pooled variance with the logs of the group variances, and is divided by a correction factor. Under the null hypothesis it follows a chi-square distribution with k − 1 degrees of freedom, which gives the p-value.

When should I not use it?

The test is very sensitive to departures from normality: a heavy tailed sample can give a small p-value even when the variances are equal. If your data are not close to normal, prefer the Levene or the Brown-Forsythe test. It also needs at least two observations per group.

What do I do after a significant result?

Unequal variances break the assumption of the classic ANOVA, so use Welch’s ANOVA, which does not assume equal variances, or transform the data. A p-value above 0.05 does not prove that the variances are equal, it only means that no difference was found with this sample.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.