On Thu, 20 Jan 2000 10:46:47 -0600, "Jason Osborne, Ph.D."
<[EMAIL PROTECTED]> wrote:
> I am testing for partial mediation. I need to know whether the
> unstandardized regression coefficient for variable X predicting Y in one
> regression equation is significantly different from the unstandardized
> regression coefficient for variable X predicting Y after a couple
> covariates have been entered, in a seperate regression equation.
- non-nested hypotheses are mentioned in my stats-FAQ.
(Why do you say "unstandardized" when that is the one everyone would
imagine looking at, anyway? You almost threw me off, with that one.)
("Standardized" -- within one sample, it could not make any
difference, since you would be dividing by the same SDs to
standardize. Between samples -- who could want standardized?)
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
