Cochran's test

Cochran's test

[′käk·rənz ‚test] (statistics) A test used when one estimated variance appears to be very much larger than the remainder of the estimated variances; based on the ratio of the largest estimate of the variance to the total of all the estimates.