finding p-value when F-value and df's are known

[hanelliot]

Websites like this: http://www.danielsoper.com/statcalc/calc07.aspx
gives you exact p-value when u enter F-value, and df's. Is there any way we can obtain this from a F-table only? I can manage to get the range, but not the exact value. This is normal?

[gianmarco]

Hi!!

Provided that any statpack gives you the p value for F-test, you could try to use Excel as I pointed out to you in my reply to a previous post of yours about chi-square (http://www.talkstats.com/showthread.php?t=11059).

In Excel:
1) click on any cell

2) type the following formula:
=distrib.f(...)

3) instead of the dots, put the F value and the degrees of freedom of your two samples, each separated by ";"

Excel should return the associated probability.

Please note that you have to put first the DF of the sample going in the numerator (that is the sample with the greater variance), and then the DF of the sample going into the denominator (that is the sample with the smallest variance).

Hope this helps
Regards
Gm

[hanelliot]

Hi, is there any way I can do it without using excel? We are talking about a test environment.. only F-table available. Thx

[TheEcologist]

Hi, is there any way I can do it without using excel? We are talking about a test environment.. only F-table available. Thx

Yes there is, check this post first:
http://www.talkstats.com/showpost.ph...75&postcount=6

It wont be easy if your not a maths person though.
Heres the page for the F-distribution.
http://en.wikipedia.org/wiki/F-distribution