F-test

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

F-test
Message posted by Michael Chinn on July 25, 2000 at 12:00 AM (ET)

What is a F-Test and how do you know when the final numbers on the F-test is significant. In other words, what do the numbers mean?

READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: F-test
Message posted by nancy diehl on July 25, 2000 at 12:00 AM (ET)

An F-test compares the variances of two sets of data to see if they come from the same population variance.
It is often used prior to the t-test on the means so you know whether the variances are equal or unequal.
The idea of the test is to create a ratio of the larger variance divided by the smaller variance. The closer
this ratio is to 1, obviously the variances are then similar. However, this test is also a function of the sample
size so you need to take the degrees of freedom (df) of the larger sample variance and the df of the smaller
variance sample and look in the F-table for a comparison number based on these degrees of freedom. If your
calculated ratio value is larger than the value found in the F-table, then you reject the null hypothesis
and conclude that the variances are different and that the two samples come from different distributions.

Your $5 contribution helps cover part the $500 annual cost of keeping this site online.