HomeRoast Digest


Topic: OT now: Re: Awfully small sample (32 lines)
1) From: Jim Gundlach
Paul,
     The study focuses on mental alertness measured by performance on  
a complex word puzzle.  The hypertension finding was a side finding  
that they went ahead and reported because it might be important.   By  
using the controlled experiment, they were controlling for the other  
factors.
     The Harvard Physician's Heath Study was designed to look for all  
kinds of factors that impacted health.  They measured hundreds of  
variables several times over more than two decades.  The aspirin study  
was done when someone wanted to look at the effect of aspirin and  
discovered that the Harvard Physician's Health Study already had data  
that would let them look at the question.  They used the available  
data to produce immediate results economically rather than doing an  
entirely new study that would be expensive and take a long time.   
Given the strength of the aspirin effect, it would not have required  
anywhere near 20,000 cases to show that the effect was significant.     
They chose to look at 20,000 cases because they wanted to be able to  
detect very small effects on health when they started the study.   
Given that they were looking for things that might cause people to die  
early, they decided that it was important enough to not have small  
differences statistically insignificant.
    pecan jim
On Aug 27, 2008, at 11:54 AM, Paul Helbert wrote:
<Snip>
Homeroast mailing list
Homeroasthttp://www.sweetmariascoffee.com/gallery/main.php?g2_itemIdx20">http://lists.sweetmariascoffee.com/listinfo.cgi/homeroast-sweetmariascoffee.comHomeroast community pictures -upload yours!) :http://www.sweetmariascoffee.com/gallery/main.php?g2_itemIdx20


HomeRoast Digest