# Thread: Testing equality of variances using the F distribution.

1. ## Testing equality of variances using the F distribution.

I have a question where I have to test the equality of two variances using the F distribution at 5% level of significance.

H0 is the variances are equal
H1 is the variances are not equal.

I woked my p-value out to be 1.19 and I have to use an F29,29 distribution.

Do I have to use the 2.5 percent points table as it is a two-tailed test?

Also if 1.19<F29,29 do I accept or reject H0?

Can anyone help me?

2. Originally Posted by Nic
I have a question where I have to test the equality of two variances using the F distribution at 5% level of significance.

H0 is the variances are equal
H1 is the variances are not equal.

I woked my p-value out to be 1.19 and I have to use an F29,29 distribution.

Do I have to use the 2.5 percent points table as it is a two-tailed test? Mr F says: Yes.

You reject H0 if the calculated F-value is in either the the upper or lower 2.5 percent of the tail of the F distribution. The upper-tail critical value can be got directly from the tables. Do you know how to get the lower-tail critical value?

Also if 1.19<F29,29 do I accept or reject H0? Mr F says: See above.

Can anyone help me?
..

3. Thank-you.

Is the lower tail critical value 1-upper critical value?

4. Originally Posted by Nic
Thank-you.

Is the lower tail critical value 1-upper critical value?
No. The distribution is not symmertic in that way.

Try doing some research on it and if you're stuck post again in this thread.