linear relationships

[FIELDRW@aol.com]

In a message dated 3/8/00 3:04:26 PM Central Standard Time, tdc@xrayted.com 
writes:

<< Here's a stupid question - when it comes to experimental design and
 testing validity.
 
 Following this thread I see discussions of confounding factors that are
 difficult to control for because they are non-linear.
 
 Well the question is this:
 
 If the end point is "X" and there are multiple possible causal factors
 AND from the end point "X" it is not possible in any way to discern
 WHICH insult was the actual causal factor - HOW is it reasonable to
 accept that the influence of one agent is linear with exposure - and
 another is not? >>

Ted, That is not a dumb question.  Please email me directly about this, it 
may take a few emails to explain it.

Regards, Bill

mailto:bill-field@uiowa  
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html