|
Hello
Recently I posted a message to ask whether it is possible to run a chi-square test of linear trend in conjunction with a logistic regression analysis (and hence so as to control for confounding factors). I was provided with the following code (within which LevelOfExposure is the name of my independent variable):
use /CONTRAST (LevelOfExposure)=Polynomial
and the following advice on this code
"It will decompose LevelOfExposure in k-1 (k:Nr. of levels) variables.
The first one will test the linear trend, the second one the quadratic
term, and so on."
Whilst I am grateful to have received this info, it would also be most useful
if someone would kindly point me to any literature which explains this version of the
chi-square test for linear trend in detail. Also, I would be interested to
learn whether it would be equally appropriate to run such code in
conjunction with a Cox regression analysis so as to perform a chi-square test
of linear trend for mortality (with controlling for confounding).
As I am not familiar with the above code, it seems far
from obvious to me why it would in and of itself
control for the other factors in the model. Some advice here
would also be valuable.
Many thanks in advance for this assistance.
Regards
Margaret
---------------------------------
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine
|