Re: More Questions about analysing ABMs

[Docgca]

Steve,

Thanks so much for your comments.  I would really appreciate it if you could 
send me that paper you mentioned.  Now for some responses to your comments:

You wrote: I always avoid comparing the distribution of results from multiple 
model
 replicates (varying only in random number seed) to distributions of
 observed data- for this comparison to be valid, your clinical data would
 have to be collected from very very similar injuries. When you compare
 distributions of model results to data, you have to make sure the same
 process was causing the variation in both cases. Rarely in nature do
 "experiments" have the same initial conditions like replicate model runs
 do.

I guess what I'm trying to do with the RNGs is to model the heterogeniety of 
the patient population/response to a fixed insult.  The problem with clinical 
studies is that there is too much difference in the initial conditions (read: 
patients entered into the study) so that they have a lot of trouble getting 
enough "n" to see if there is any statistical difference.  As a result a lot of 
these studies show "no statistical significant results," bue do have 
"tendencies." That's a little different than what I'm tryiing to do in this 
phase of model testing with the same initial injury and the RNGs on:  I think 
this is determining the "noise" that comes from the RNGs.  Am I wrong in this 
assumption?  In my paper I showed a graph of a "representative run." The 
reviewers had asked for some determination that the "representative" run I 
displayed was not something I specifically picked and/or was an outlyer.  I had 
hoped to show that the run I picked fell in the peak of the distribution !
at a single initial injury number, and it does.

You wrote:
 Some things to worry about are that the statistical
 significance of the difference between scenarios (with, without
 intervention) depends on the number of model runs. Even if your
 intervention has only a very small effect, it will be a statistically
 significant one if you compare results of 1000 model runs. So be sure to
 examine the medical significance as well as the statistical
 significance. 

This gets back to the purpose of the simulation.  Does a very high "n' mean 
that you can prove anything statistically significant?  That's actually not too 
bad, since a lot of these interventions actually hurt more than they help; if 
that could be identified prior to trying it on people it might be nice to know.

Thanks again for your comments, and I look forward to your response.  And if 
you could send me that paper I would really appreciate it.

Yours,
Gary An, MD
Department of Trauma, Cook County Hospital
Chicago, IL


                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================