Re: Get your tickets now - New Discussion List

[Jerry, Cohen]

To help those of us who may not be among the cognoscenti, could someone 

please explain the concept of  " unbounded residual confounding ". Could this phenomenon also account for the apparently positive association between 

smoking and cancer?





----- Original Message ----- 

From: <niton@mchsi.com>

To: BERNARD L COHEN <blc+@PITT.EDU>

Cc: 'Radiation Safety' <radsafe@list.vanderbilt.edu>; <RADHEALTH@LIST.UIOWA.EDU>

Sent: Friday, July 04, 2003 11:25 AM

Subject: Get your tickets now - New Discussion List





> Dr. Cohen,

> 

> Many people have offered suggestions to help you with your plea to explain 

> your inverse associations we all see with your ecologic data.  

> 

> As Edward Teller sated, Two paradoxes are better than one; they may even 

> suggest a solution. 

> 

> I believe Dr. Teller's statement is very applicable to our discussion.   

> 

> Paradox 1: Your well known inverse association between smoking and lung cancer.

> 

> Paradox 2: The less than well known inverse association between your county 

> radon data and other smoking related cancers.

> 

> The existence of these two paradoxes helps to explain that the cause of these 

> inverse associations are merely unbounded residual confounding from smoking  and other within county variables associated with smoking