Ventana software support forum

Hi,

Does anyone know the difference between Sterman's approach and Soft Systems Methodology? If needed, how would I use SSM to make Sterman's approach more robust?

Dave

Hi

A model is more robust if the risk of its results being far from reality is minimal.
The difference of Sterman's approach and SSM relatively to the results will depend on the problem, the ressources allocated to the building of the model: competence of the modeller, time to build the model, quality of the people having to implement the solution etc..
To understand the theoretical difference of both techiques one must buy some books.
To have a short description of SSM, buy the book 'Rational analysis for a problematic world revisited' from Jonathan Rosenhead from Wiley. This book exposes too other methods; Soda (decision explorer software), Strategic choice approach (Strad software), drama theory and Robustness analysis. All these methods are worth being studied because they take different approaches to problem solving and are more or less useful depending on the kind of problem. Other methods like SD, VSM and MCDA are too very shortly described. A chapter is too dedicated to multi methodology. By example SSM like Soda (both are more systemic methods like VSM) are often used with SD and can evntually make Sterman's approach more robust.
Regards.
JJ

This question is answered very nicely in a book that I will continue to reference.

Creative Holism for Managers

This book was invaluable for me.

As I see it (my opinion), SSM is designed to be helpful in areas where people have different perspectives, purposes, and problem definitions. Sterman's approach does seem to address soft issues, but I think that Sterman overlooks or doesn't address some issues such as "defining problems in areas where there are power imbalances and language issues that are real aspects of some actual organizations".

Sterman is an excellent System Dynamics modeller. I've got his Business Dynamics text book, and I find mapping out soft variables isn't really that hard, haha. Use what you think works best for your problem.

Certainly, hard variables are easier to map out as there is a lot less uncertainty. Heck, even for some softer unknowns in a hard model, statistical data can give pretty good values for your ratios in your equations.

"Soft Systems Methodology" from Wiki:
1. Appreciating the unstructured problematical situation
2. Understanding the worldviews of the key stakeholders
3. Creating root definitions of relevant systems
4. Making and testing conceptual models based upon worldviews
5. Comparing conceptual models with reality
6. Identifying feasible and desirable changes
7. Acting to improve the problem situation

From Sterman's Book, Business Dynamics (pg 852)
1. Purpose, Suitability, and Boundary
2. Physical and Decision-Making Structure
3. Robustness and Sensitivity to Alternative Assumptions
4. Pragmatics and Politics of Model Use

At an iSee conference, an the model facilitator for an oil pricing model used Checkland's approach (looking at "messes") to formulate causal loops, and THEN turned those loops into stocks and flows, and simulated the outcomes.