Re: dose limits for members of the public

[AndrewsJP@AOL.COM]

In a message dated 2/14/2002 6:19:03 PM Eastern Standard Time, sandyfl@EARTHLINK.NET writes:



> THEREFORE, cars are 
>
> safe and do not injure people ... by reason of my personal experience.


Mine is, the other guy's is not.


I keep trying to not respond this flippant way.  It wastes bandwidth.  But I just had to.


Here is another thought for Ruth W.  and her transportation analyses of risk.  I note that there is not much mention of the error bars associated with this kind of analysis and it can be done with current software.  Ruth, is that being done routinely with your analyses?  Do you assign a distribution and range of values for the error term of values that you use in these calculations and determine the probable distribution of results?  If you do that, at what level of error distribution do you say "We just don't know!"


Several years ago I got some developing software of this type and the results were very interesting.  One of the results reported at that meeting was the huge error bands that result from duplicating some of the EPA computations on risk.  The error bounds for the 95% CL were 5 to 6 orders of magnitude between the upper and lower bounds.  This is clearly in the "I don't know" region.


Ruth, you don't have to answer the question, but it would be interesting for anyone who has current experience using this kind of program for risk analysis to give us his or her thoughts.


John Andrews

Knoxville, Tennessee