Question

If you are comparing subgroups of a data set, how can you determine if the variance between subgroups is significant?

Answer 1

hypothesis testing using `chi^2`.

Explanation:

hypothesis testing using `chi^2`.

The null hypothesis is `sigma_1^2 = sigma_2^2`
The alternate is `sigma_1^2 != sigma_2^2`

The test statistic is `t=sigma_1^2/sigma_2^2`. The more we deviate from 1 the more likely we are to reject the null.