In article <[EMAIL PROTECTED]>,
Jason Osborne, Ph.D. <[EMAIL PROTECTED]> wrote:
>This is a multi-part message in MIME format.
>--------------BB4FA7270D1DE5E3C7A6342A
>Content-Type: text/plain; charset=us-ascii
>Content-Transfer-Encoding: 7bit
>I am trying to test a mediational model via multiple regression, to see
>of the beta for the relationship between X and Y is significantly
>reduced by covarying A. I need to test for significant differences
>between the beta X -> Y without A in the equation, and with A in the
>equation.
>These don't strike me as independent betas. Are they? I have skimmed
>through several regression texts, such as Cohen & Cohen and Pedhazur, as
>well as reading baron & Kenny's 1986 J Personality and SSocial Psych
>article. I cannot find a formula for testing differences between a
>particular regression weight in successive regression models.
It is not clear from what you wrote if X or Y is the independent
variable; I will assume that the regression is of Y on X; X may even
be multivariate, as may A, in which case the analysis is a little
harder.
In the case X and A are single variables, the true regression will
change by including A unless the true regression on both does not
involve A, or X and A are uncorrelated. In all other cases,
including A changes the coefficient of X.
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
[EMAIL PROTECTED] Phone: (765)494-6054 FAX: (765)494-0558
