There's an interesting but sad article in the NY Times concerning
failed cancer research coming from Duke researchers use of a
genetic analysis for detecting cancer and for treatment; see:
http://www.nytimes.com/2011/07/08/health/research/08genes.html?_r=1&nl=todaysheadlines&emc=tha2

One of the main points is that the amount of data that is collected
in this approach is huge and the likelihood that there are significant
errors in the data have been dismissed as "clerical errors" that
had no effect on their conclusions.  When a couple of statisticians
examined the data from one project, they uncovered a variety of
errors in the data and in the conclusions reached.  The researchers
appeared to believe that they had "gotten it right" and proceeded
to build research empires on this quicksand.  This belief or
"intuition" was commented on by one of the statisticians:

|“Our intuition is pretty darn poor,” Dr. Baggerly said. 

I have a feeling we might find something similar in the neuroimagining
area.

-Mike Palij
New York University
[email protected]







---
You are currently subscribed to tips as: [email protected].
To unsubscribe click here: 
http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=11325
or send a blank email to 
leave-11325-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu

Reply via email to