QUERY Policy paradox and SD (SD6661)

[SDMAIL Monte Kietpawpan]

Posted by  Monte Kietpawpan <kietpawpan at yahoo.com>

Dear all,
   
 Structural confirmation is a primary test of model validity. 
If model components and equations can be easily 
recognized, the model would pass this test. This is not 
a strong test. A big problem occurs when policies are 
easily recognized but somehow impractical. All policies 
can be implemented in SD models with ease, but many 
of them may not be easily implemented in the real system. 
We know what policies can improve the system performance, 
but such policies do not exist. SD practice per see is not 
consistent with the real system. 
   
SD paradigm underestimated difficulty in parameter 
adjustment in the real world. Many problems keep persist, 
not because people have implemented the wrong policies, 
but because they cannot implement the right ones, known as 
"policy paradox".
   
SD modelers represent policies as exogenous factors, 
while in fact they are edogenous. We are teached to view 
ourselves as a plane designer, but the system we try to 
represent is usually not a plane or an object under our control. 
You can modify your own plane as much as you want, but you 
cannot modify the social system, becuase it is not yours. 
  
Monte Kietpawpan
PSU
Posted by  Monte Kietpawpan <kietpawpan at yahoo.com>
posting date  Sun, 9 Dec 2007 20:39:01 -0800 (PST)