[ RadSafe ] Radon: POWERFULLY associated with LESS lung cancer by B.Cohen

[Bernard L. Cohen]

On 6/17/2011 5:17 PM, Brennan, Mike (DOH) wrote:
> Dear Prof. Cohen and Prof. Raabe,
>
> I assume that you are familiar with Dr. Field's Iowa Radon Lung Cancer
> Study.  I would be interested in your thoughts on it, and if he was
> faulty in his methodology, what that fault was.
         --I am not an expert in the technology of case-control studies 
so I am not qualified to offer publishable criticism. Case-control 
studies obviously have problems with limited statistics and I distrust 
pushing very deep into statistical theory to filter out results -- the 
assumptions going into statistical theory are imperfectly fulfilled in 
such studies. Their adjusting for potential confounding factors by use 
of multiple regression has serious limitations which I have 
demonstrated. Their matching between case and control groups is far from 
perfect in cases I have looked into -- they are often matched by age but 
not by economic status, education, ethnicity, and other such potential 
confounding factors which can be very important.
         --My studies have treated over 500 such potential confounding 
factors and shown that they do not change the results. My treatment of 
confounding factors is much more sophisticated and accurate than their 
multiple regression. Statistical fluctuations play no role in my 
results. No one has ever offered a criticism of my studies that I have 
not shown to be very clearly invalid. If anyone does not agree on that 
point, please let us discuss the matter.

-- 
Bernard L. Cohen
Physics Dept., University of Pittsburgh
Pittsburgh, PA 15260
Tel: (412)624-9245  Fax: (412)624-9163
e-mail: blc at pitt.edu  web site: http://www.phyast.pitt.edu/~blc