AW: [ RadSafe ] Re: Residential radon risk

[Bernard Cohen]

Franz Schönhofer wrote:

>"Why do you ignore my response.....", just a copied phrase from your mail.
>Do you believe that it is below your scientific level to answer my clearly
>formulated criticism? Why do you still mention average county radon levels
>in this post? 
>
          ---Your message was not addressed to me specifically and was 
not about my work specifically, so I did not feel obligated to respond, 
butI will respond here.
    The health effects of radon depend basically on three things, radon 
gas concentration -r, concentration of radon daughters --WL (working 
level), and unattached fraction -UF (fraction of radon daughters not 
attached to a dust particle - this is important because these have a 
much greater probability of sticking to the bronchial surfaces). In 
mines, there is so much dust that one can assume UF = 0, so measuring WL 
gives the health effects. In homes, this is not so. As an example, one 
can drastically reduce the WL by removing the dust (for example, with an 
electrostatic precipitator), but the newly formed radon daughters have 
no dust to attach to, so UF = 1.0, and the danger is not reduced. 
Roughly, health effects, HE = k x WL x UF.where k is a constant
     In general, there is a strong negative correlation between UF and 
the ratio of WL/r (equilibrium factor), or .UF x WL/r  = K, another 
constant.  Combining these two equations, HE = k x K x r, or health 
effects are proportional to r, HE = k' r. Detailed studies have 
confirmed this result, and shown that it is much more accurate than 
assuming HE = k'' x WL. (k' and k'' are new constants)
    Ideally, one should measure WL and UF, but that is very difficult 
and is essentially never done. Moreover, it is much easier to measure r 
than to measure WL. That is why everyone measures r. .
    In my studies involving hundreds of thousands of measurements of r, 
it seems reasonable to assume that there is no strong systematic 
variation in the ratio of r to health effects. If you have reason to 
think that there might be such a systematic variation, please let me know. 

>I insist that it is more than unscientific to claim minor statistical
>significance, when the data might be wrong by tens of percents.
>
>Sorry to say, that I have once admired your way of proofing that others are
>incorrect, by using their data and showing that they are in disagreement
>with the claimed results. I have used this tactic very often myself. 
>
>In this case I believe that you are working with data (radon
>concentrations), which are not directly related to your "results" - lung
>cancer. 
>
>Franz
>
>
>  
>
>  
>