Number12, May 2010
A COMPENDIUM OF MODELLING TECHNIQUES
Available online at: http://i2s.anu.edu.au/sites/default/files/integration-insights/integration-insight_12.pdf
Thank you to the BSSR listserv for making me aware of this resource.
This resource nicely allows all of us who call ourselves “complexity thinkers” to hopefully understand in general what other languages are raving about. With a quick, broad stroke, Badham highlights “three” overarching categories under complexity:
1) Qualitative aggregate methods (including Churchland’s soft systems methodology that I desperately want to learn, hopefully from the master, and methods such as concept maps and causal loops that I had already used in my life as a qualitative researcher) but honestly how much do typical complexity thinkers given credence about qualitative when it screams an unreliable mental model to them?
2) Quantitative aggregate methods such as Bayesian Nets, Dynamical Systems, Systems Dynamics, as well as “regression” which come to mind to most when we evoke the term “complexity”- But I must be honest I did not feel at all like a complexity thinker when I analyzed data to the world of regression. “Complexity” requires a macro application and discussion beyond Micro residual plots and should ideally take the discussion beyond the SAS printout. But this has not happened in any statistics course that I have taken…Was it even the place for that discussion? If not, then when?
3) Individual Oriented models such as cellular automata, agent based models, and social network analysis (YES!!!!!) .
So I am a “1” and “3”, and not a “2”. But there is room in complexity for all of our voices, right?
It is a quick read with illustrations and additional references to consult.
Personally, when I realized that I evaluated the world using “complexity”, it was an arduous process to accept that fact and THEN find and hone a method that fit my interests as well as my innate talents and skills. The table of complexity is long and wide and handling the uncertainty in the world needs all three….and I could say that perhaps a little more parity across categories at this table should be a goal of our community.