View Single Post
Old 02-25-2009, 03:55 PM   #10
tedrodgerscpa
Palmetto State Herf Crew!
 
tedrodgerscpa's Avatar
 
Join Date: Oct 2008
First Name: Ted
Location: Atlanta
Posts: 1,712
Trading: (87)
QdO
tedrodgerscpa has disabled reputation
Default Re: stirring the pot local B&M vs big internet

Quote:
Originally Posted by WildBlueSooner View Post
I propose this. Take the same stick and give people 4 of the same stick. However they will get them in any combination...could be 2 from online and 2 from B&M, one and three, or 4 from one source and have them rate them and say which they think is which. I can almost guarantee you there will be no trend in either direction.
Quote:
Originally Posted by SeanGAR View Post
I don't know how to run the statistics on that type of experimental design. Anybody???
Quote:
Originally Posted by taltos View Post
Can't see any control group there and don't see much experimental design.
That looks to be a triple blind study, which is the strongest test in terms of causality, and is rarely used.

If I wanted to hypothesize that real bullets are deadly and blank bullets are not, I put a mess of bullets in a box. I don't know which bullet is real, and which is blank. Put gun to head, pull trigger.

If I die, I have unmistakeable proof that real bullets are lethal.

However, if I live, I don't necessarily have unmistakeable proof that bullets are lethal, as I may have drawn a blank.
__________________

Everyday I'm shufflin...

Last edited by tedrodgerscpa; 02-25-2009 at 03:56 PM. Reason: Clarification
tedrodgerscpa is offline