On Fri, 8 Jun 2001, Mike Tonkovich wrote:
> I ran an analysis of covariance using PROC GLM (in SAS) with an
> interaction statement. My understanding was that a nonsignificant
> interaction term meant that the slopes were the same,
Not exactly. It means only that differences in the slopes, if such
differences exist, were not large enough to be detected in the context
of your analysis. As another respondent has observed, there may be
issues of power to contend with here.
> and vice versa for a significant interaction term.
Well, not exactly "vice versa". A significant interaction term would
imply that differences in slopes WERE large enough to be detected (or,
possibly, that the differences observed constitute a Type I error).
> Is this ... the best way to approach this problem with SAS?
Can't tell, because you haven't described "this problem".
Another way to approach the problem, of course, is to think of
it as a multiple regression problem with (I presume) some categorical
predictors. Whether this would be a "better" way depends partly on the
output SAS produces for regression analysis and the output it produces
for analysis of covariance (I am not a SAS user, so cannot offer advice
on this point); and partly on what "this problem" really is.
> Any help would certainly be appreciated.
I hope this has been helpful.
-- DFB.
------------------------------------------------------------------------
Donald F. Burrill [EMAIL PROTECTED]
184 Nashua Road, Bedford, NH 03110 603-471-7128
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
