The independent variables don't need to be normally distributed. If you
log transfom the variable than you think that if you start with a small
dose, an increase has a much bigger effect than when you start with a
larger, which may or may not make sense. Anyhow the real issue with
transforming the indepedent variables is whether the effect is linear,
or gradually slowing down (log) u-shaped, quadratic, etc. As an aside,
if you compute percentile rank scores, you won't make that variable
normal, uniformly distributed.
-- Maarten
--- Mohammed El Faramawi <[email protected]> wrote:
> I have questions about the independent variable
> transformation.
> I have an independent variable (lead mg/dl) which is
> not normally distributed. I am trying to run it in a
> cox hazard model the out come is mortality. What
> should I do with lead
> 1) Should I log transform it? I did that but it is
> still not normally ditributed after transformatin.
> 2) Leave it as it is because transformation does not
> change the distribution?
> 3) categorize using quartiles or percentiles?
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
___________________________________________________________
Rise to the challenge for Sport Relief with Yahoo! For Good
http://uk.promotions.yahoo.com/forgood/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/