[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Cancer deficiency clusters



This reason is exactly why statistics profs advise

people doing _research_ to use 2-sided confidence

limits, and discourage the practice of 1-sided limits

in most cases.  [1-sided limits have some valid

applications in non-research fields such as quality

control and business management].  



Along with this recommendation came the recommendation

not only to look for what one 'thinks might happen' in

an experiment, but to look for _anything_ that's

happening.



Fortunately, the NSWS used 2-sided limits, although

there are studies by less competent epidememiologists

that used 1-sided limits (improperly IMHO).



~Ruth 2  



--- Ted Rockwell <tedrock@CPCUG.ORG> wrote:

> > That's what I've been taught in "Research

> Methodology" cpurse few years

> ago.

> 

> Let me get this straight, guys.  Suppose I read

> about a restaurant that is

> decorated in blue.  And I think, "They've got it

> wrong.  I've been there.

> It's red."  So I go back with the research intention

> of proving that the

> restaurant is red.  And I find it's blue.  I take

> pictures and they come out

> blue,  But that doesn't proved anything, because if

> my INTENTION was to

> prove it red, then my photos can only "suggest" that

> it's blue.

> 

> C'mon.  Does anyone really believe that?  I'm

> willing to admit that an

> experiment designed to prove one thing might not be

> optimum to prove

> another. When you're intent on counting dead mice,

> you might not notice the

> happy mice.  For example, there's a famous

> experiment which reported "Since

> there is no reason to expect that radiation will

> beneficial, any deviations

> from norm will be assumed to be detrimental."  But

> that's just bum science.

> 

> What am I missing here?

> 

> Ted Rockwell





__________________________________________________

Do You Yahoo!?

Yahoo! Health - Feel better, live better

http://health.yahoo.com

************************************************************************

You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.

You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/