christophe tourenq wrote:
> Dear colleagues,
> I think that I read or heard somewhere, that in a multiple regression
> procedure, the absence of effects of some variables may be due to the
> strong effect of a few selected variables that "masks" the effects of
> other variables.
> Did you read or heard something like that or it is just a dream?
> If you read it, references would be appreciated,
Christophe:
It is certainly possible that a variable that has a sizeable
correlation with the dependent can have a weak and nonsignificant
effect in a multiple regression with other predictor variables. If
all of the predictors are forced into the equation on a single step
then you are assessing their unique effect on the dependent variable
while controlling for the other predictors. This is why one is
generally in a better position to interpret the results when the
variables are entered in a theoretically meaningful order determined
by the user. Order certainly makes a difference, and if everything is
entered at once it is as if each variable was entered last.
You may want to start with an introductory chapter on multiple
regression and then work up to a more comprehensive treatment like:
Pedhazur, E. (1982). Multiple regression in behavioral research.
hope this helps,
Chuck
----------------------------------------------
Chuck Cleland
Institute for the Study of Child Development
UMDNJ-Robert Wood Johnson Medical School
97 Paterson Street
New Brunswick, NJ 08903
phone: (732) 235-7699
fax: (732) 235-6189
http://www2.umdnj.edu/iscdweb/
----------------------------------------------
===========================================================================
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/
===========================================================================
