Re: [tlug] How much of radiations measured in Central Tokyo?

[Stuart Luppescu <slu@example.com>]

On Tue, 2011-03-15 at 16:46 +0100, Josh Glover wrote:
> Out of 3,877 crew, 169
> developed cancer, compared to 153 in a similar-sized sample of
> non-pilots."
> 
> I'm no statistician, but I wonder if the sample size if large enough
> for the result to be significant. 

I'm a statistician, but I'm no good at these kinds of problems. (I'm
actually a psychometrician, which is a different area of statistics.)
However, I'll give it a shot. It may be completely wrong.
The risk of cancer in the general population is 3.95%; the risk for
pilots is 4.36%. The standard error of the estimate would be 0.312%, so
the standardized difference would be 1.32, which is not significant at
the 5% level. 

Regardless of statistical significance, the incremental risk is only
0.41 percentage points, which is not a whole lot. In percentage terms,
that means about a 10% increase in risk, but since the probability of
contracting cancer in the first place is pretty low, the increase in
risk is not much. Put in other terms, an additional 0.3% of the
population is likely to get cancer resulting from the exposure.
-- 
Stuart Luppescu -=- slu .at. ccsr.uchicago.edu        
University of Chicago -=- CCSR 
才文と智奈美の父 -=-    Kernel 2.6.36-gentoo-r5                
So good advice here is: Beware of good advice about
 this. (Of course, I may just be an outlier ...)   
 -- Berton Gunter (replying to the question what
 the best way to detect an       outlier is)       
 R-help (September 2004)

Attachment: smime.p7s
Description: S/MIME cryptographic signature