Re: Cohen's Ecologic Studies

[Ted de Castro <tdc@XRAYTED.COM>]

Excuse me - but - it seems to me that analysis using aggregate data MUST

presuppose that effects are linear and you are using aggregate data for

analysis of confounders.



If response to confounders is NOT linear - then they can be accounted

for only by groupings in discrete exposure level groups and NOT ranges

and such grouping would require individual data.



I am not an epidemiologist - but then this notion isn't epidemiology -

its just plain arithmetic.



BERNARD L COHEN wrote:

> 

> On Mon, 3 Jun 2002 EPIRAD@mchsi.com wrote:

> 

> > The bottom line as I have been trying to get across in

> > HP publications for years is that smoking is not a

> > linear function within or across counties in the United

> > States.  If it were, then you could possibly adjust for

> > it better at the ecologic level.

> 

>         --My several independent treatments of smoking make no assumptions

> about smoking being a linear function of anything

> 

>  As for Dr. Cohen's

> > ecologic analyses, he can account for only about 30% of

> > the lung cancer mortality with his smoking data.

> 

>         --Wrong, wrong, wrong. The fact that R-squared is only 30% derives

> from the small up and down statistical variations. The true indication of

> predictability is the standard deviation of the slope of the regression of

> lung cancer on smoking prevalence which is very small percentage-wise

> 

>  Cohen's

> > derived LNTT formula does not have the ability to adjust

> > for non-linear covariates like the BEIR VI formula

> > does.

> 

>         --Wrong, wrong, wrong. My treatments of confounding factors do not

> depend on linearity. My treatment was based on BEIR-IV as it was the

> latest thing available. I could do a treatment based on BEIR-VI if it

> would be publishable, but I doubt if it would be considered new enough to

> be published

> 

> ************************************************************************

> You are currently subscribed to the Radsafe mailing list. To unsubscribe,

> send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

> radsafe" (no quote marks) in the body of the e-mail, with no subject line.

> You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/

************************************************************************

You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.

You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/