On Fri, 17 Aug 2001 13:28:21 -0400, Paige Miller
<[EMAIL PROTECTED]> wrote:
> Elliot Cramer wrote:
> >
[ snip, question ]
> >
> > The test of parallelism in mancova is an equivalent test
I can believe someone would call it that, but I can believe other
tests would be called that, too. You state this is, in reference
to what test (what computer program, or what textbook)?
>
> I'd like to ask a follow-up question then. MANCOVA uses least squares as
> its objective function to estimate relationships, while canonical
> correlation uses a different objective function. They don't seem
> equivalent to me, so
[ break ]
MANOVA is a canonical correlation problem. MANCOVA, so far
as I imagine it, is another canonical correlation problem. You might
be taking steps in a particular order, but you are still solving
to find the eigenvalues of an asymmetric matrix -- at least,
most of the time.
Is this "all the time"? - it probably depends on your universe
of problems. The two-group comparison described in this
thread was an eigen-problem.
> my question is: is there some math that I'm not
> aware of that shows these two are equivalent? If so, could you provide a
> reference?
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
