REPLY Society Strategy Development (SD6938)

[SDMAIL Brian Crowe]

Posted by  Brian Crowe <brian_crowe at i-worx.com>

I might use SD to help answer questions from my clients concerning the 
repercussions of my recommendations.  Two areas that push significant 
complexity into a SD model concern human learning systems and human 
decision making systems.  In these areas, I defer to conversations with 
my client in lieu of a highly complex and unvalidated model that I am 
not expert enough to create and demonstrate credibly.  It would be 
useful to have a "clearing house" of validated models to draw upon - 
this requires something called a "description language".  There has been 
some discussion on this here already, and I think that  it would be 
something valuable to work on in the SD Society.

I appreciate the follow-on remarks from Bob and Doug.  It would be great 
if CNN reported that scientists applying principles of System Dynamics 
have shaped sweeping improvements in K-12 education systems worldwide, 
for example.  I am still "brain-storming" here, and would like to build 
upon Bob and Doug's suggestions to draw on other fields of science to 
augment SD work in the two problem areas I noted above.

Human learning is not a linear system - it is step-wise in character as 
the student encounters plateaus of understanding and experiences 
epiphanies and reaches the limits of teaching resources and their own 
abilities.  Human learning is incomplete, and that has consequences.  
And knowledge is forgotten and must be refreshed...  In the US, there is 
no tailored feedback loop in K-12; a missed problem on an exam rarely 
prompts a return to fundamental principles.  It would be interesting if 
an education crusader waved their SD model in an official's face that 
demonstrates the dramatic impact from the addition of a new feedback 
loop.   I am sure that much of this work has been done, but it has not 
effectively made its way to a "customer" - Doug, thanks for your 
comments here regarding the value of best practices in other fields - we 
may need to look to outside experts, and Juran would have had something 
to say regarding K-12 education quality improvement.

Human decision making instances in a population appear as a skewed bell 
shape on a time line, and it may be skewed either side of a mean time, 
and there may be several peaks of decision making activity over time.  
The emergence of consensus has a life cycle.  Early decisions are often 
the result of exuberance and late decisions are often the result of fear 
of consequences (both factors are elements of risk aversion).  Apathy, 
prejudice and incomplete information contaminate the decision process in 
both quality and timeliness.  The Presidential campaign here in the US 
comes immediately to mind...  I would be interested to know who of the 
trio of candidates uses an SD model in their planning - I bet at least 
one of the candidates uses a SD model, but they may not recognize it for 
what it is.

These elements of social SD models are perplexing for an unsophisticated 
modeler like myself, and failure to capture nuances properly results in 
a poor model that might fail to illustrate the tendency toward certain 
irrational outcomes and behaviors that I compared to "strange 
attractors" - Bob, thanks for your comments here regarding complexity 
theory's relationship to SD.  It might be highly valuable to draw a 
complexity theorist and a psychologist into a modeling project on 
decision making processes.

A particular application of decision making that I personally find 
fascinating is human conflict.  Human conflict on any scale is often the 
result of contaminated learning and irrational decision making that 
overrides rational behavior assessments. That might become something 
very valuable to model, too.

Human systems are often linear and  "steady state" to a point.  When 
human systems are perturbed past a threshold, some element of behavior 
becomes unstable and seeks a new steady state behavior.  I frankly have 
no idea where to start modeling this aspect of behavior transition with 
SD methodologies.  Mastery of this phenomenon in an SD model would help 
answer a lot of interesting questions and solve new kinds of problems.  
There is surely significant SD development in the Human systems space 
that I am unaware of, but I also suspect that there is a lot of work 
that remains in this area.   It may be vital to draw experts and best 
practices from other fields of science into our work as often as we can 
to make progress - and to make that progress visible outside of our 
society and to an audience of potential SD users.

I am sorry to be so long winded - I tend to convey too much information 
in one "dose"...  Thanks for reading my posts.

Brian Crowe
President, TELE-WORX
Posted by  Brian Crowe <brian_crowe at i-worx.com>
posting date  Fri, 18 Apr 2008 10:08:18 -0500