It seems to me that the SAS documentation ought to tell you what algorithm
it uses to compute the degrees of freedom. Perhaps they do that by
referencing a published paper. If so, just read the original source to get
the relevant formulas. The only reliable way to get the same DF that SAS
provides is to use the same formula for calculating it. There may or may not
be existing R code that implements the required formulas, but you need to
know what you're looking for in order to find it.
Steven J. Pierce, Ph.D.
Associate Director
Center for Statistical Training & Consulting (CSTAT)
Michigan State University
E-mail: pierces1 at msu.edu
Web: http://www.cstat.msu.edu
-----Original Message-----
From: Charles Determan Jr [mailto:deter088 at umn.edu]
Sent: Friday, January 27, 2012 7:08 AM
To: John Maindonald
Cc: r-sig-mixed-models at r-project.org
Subject: Re: [R-sig-ME] Trouble Replicating Unstructured Mixed Procedure in
R
Thank you John, I truly appreciate all the advice I have received. At no
point to I feel entitled for someone to provide this for me. I do hope
that others may find this thread useful as well. If I may just as one more
question. Is there an accepted way to calculate the degrees of freedom in
this case to come to the value of 25? This has been something I have been
trying to determine.
Again, thank you to all for providing me with all the information you have,
Charles