On 18 May 2000 15:37:05 -0700, [EMAIL PROTECTED] (G.F. Gebauer)
wrote:
> Sorry to bother you with a minor "student" problem. I was referred to you
> by professor Howell who suggested you may be able to
> help me with a statistical problem I have not been able to solve myself so
> far. To put it briefly, I would like to correct multiple and canonical
> correlations for attentuation, however, I am not aware fo the correct
> formulars to carry this out. I am aware of the correction procedure for
> sinmple bivariate correlations. But I never came accross any formulars to
> correct multiple and canonical correlations for attenuation.
< ... more detail, concerning the same thing >
If it is not totally wrong in theory, I am sure that compounding
"corrected" correlations on top of each other is apt to be an error in
practice.
Why am I sure? Partly because : the business of correcting for
attenuation -- at all -- is a pretty shaky one. You have to look at
your assumptions quite carefully; you look at the circumstances quite
carefully; then you draw your conclusions. If I remember seeing it
attempted 4 times, at least 2 of them were reported in a misleading
way.
I can't say what you are comparing, but I think you want to compare
directly instead of trying to do adjustments.
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
===========================================================================
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster has no
way of controlling them, and excessive complaints will result in
termination of the list.
For information about this list, including information about the
problem of inappropriate messages and information about how to
unsubscribe, please see the web page at
http://jse.stat.ncsu.edu/
===========================================================================
