Reefer Madness?

[Stephen Swann]

On Fri, Jul 27, 2007 at 10:05:09PM +0100, Paul Mather wrote:
> On 27 Jul 2007, at 8:33 PM, vzenv14m wrote:
> 
> >Check this out as an example of manipulation of statistics.
> >
> >Kaduflyer
> >
> >http://www.msnbc.msn.com/id/19980923/
> 
> Hey, that article was pretty tame compared to the one I read earlier  
> today.  At least the one you cited didn't include the word "skunk" or  
> the phrase "modern super strength varieties" anywhere in it (which  
> seem de rigeur in such reportage right now).  (They didn't round up  
> the numbers, either.)
> 
> Still, you've got to admire the moxie of a headline writer choosing  
> to use the word "hike" in this case.  Obviously he or she figures the  
> usual gross innumeracy of the general population is a safer bet than  
> the chance of developing schizophrenia this finding represents.
> 
> It wouldn't be quite as exciting a headline if they wrote something  
> like "Boffins discover those smoking doobies increase their long-term  
> risk of developing psychosis from half a percent to a whopping almost  
> three-quarters of a percent!!!!!  (Further funding required to see if  
> the all-important one percent threshold may be crossed.)"


I noted that the researcher interviewed actually did note
that correlation does not imply causality, but the
significance of that was apparently lost on the reporter.

I'd love to a comprehensive study of the harmful effects of
alcohol, and then compare the two in terms of overall
societal damage done.

Steve