Date: Tue, 17 Apr 2012 14:08:20 -0400
Reply-To: Adam <clown_rt@HOTMAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Adam <clown_rt@HOTMAIL.COM>
Subject: MSE for a Paired T-test
When I conduct a paired t-test using proc t-test, it provides me with the
standard deviation of the differences. What I'd actually like is the MSE,
but, of course, PROC ANOVA is not appropriate for paired data.
I can't fully convince myself either way. For paired data, is the MSE equal
to the standard deviation squared, or to the (standard deviation squared)/2?
Or something entirely different?
Any advice is very much appreciated!