Date:         Wed, 17 Apr 1996 22:11:45 GMT
Reply-To:     David Nichols <nichols@SPSS.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         David Nichols <nichols@SPSS.COM>
Organization: SPSS, Inc.
Subject:      Re: (no subject given)

In article <199604171904.PAA15483@post-ofc01.srv.cis.pitt.edu>,
Michael Willoughby  <mwilloug+@PITT.EDU> wrote:
>i'm interested in calculating an exact probability level (or at least to 4
>decimal places) from the reported statistics (ANOVA) in a journal
>article--which means i have the F, df, and N available to me.  Anyone have
>any advice?
 
If you have release 6.0 or above of SPSS, COMPUTE can produce the
area under the cumulative F distribution; take 1 minus this and
you get the desired value (remember that your accuracy is limited
by how many decimals you have for the F value). The function is
 
COMPUTE PVALUE=1-CDF.F(F,DFN,DFD).
 
If you have an earlier release, you can use the FCDF function
in the MATRIX procedure.
 
--
-----------------------------------------------------------------------------
David Nichols             Senior Support Statistician              SPSS, Inc.
Phone: (312) 329-3684     Internet:  nichols@spss.com     Fax: (312) 329-3668
-----------------------------------------------------------------------------