View Single Post
  #165 (permalink)  
Old 14-May-2008, 05:43 PM
mike alexander's Avatar
mike alexander mike alexander is offline
Senior Member
 
Join Date: Jul 2003
Location: McMinnville, Oregon
Posts: 6,702
Default

Disinfo Agent wrote:
Quote:
There are no absolute rules about which value should lead one to regard a correlation as real. Statistically, any nonzero correlation can become significant, provided you gather enough data to back it up.
First sentence I agree with completely. Second I would qualify by saying that if the correlation increases as the sampled population increases you might be looking at a real causal link. If r=0.1 when n = 10, but increases to r = 0.9 at n = 100, I'm intererested. If r = 0.1 at n = 100 I'm not.

Part of the problem is that in my line of work an r of 0.5 would get me laughed out of the room.
__________________
The Devil offered me power. I told him I preferred aperture.
Reply With Quote