Community Forum : Model Analysis SIG
Welcome Guest   
 
<< first < Prev 1 2 Next > last >>
 Subject : Re:Complexity of finding all (major) feedback loops in a model.. 06/01/2019 10:18:50 AM 
Steven Wallis
Posts: 7
Location
JJ,

Fair enough as far as it goes. When working with clients, I too try to keep things as simple as possible. But I also strive to stretch the clients' thinking - to improve their systems literacy - you might say. The point of this exploration, however, is to go beyond our own assumptions so that we may create models that are of still greater usefulness and benefit to all. Or, from another perspective, we create models to better understand our real world systems; so, we'd better make models that are better. Otherwise, we are simply exchanging one set of assumptions for another (with no guarantee that either is an improvement). So, as per Coyle, I'm striving to improve our understanding of how to build better models.

Thanks,

Steve
 Subject : Re:Re:Complexity of finding all (major) feedback loops in a model.. 06/03/2019 01:49:05 PM 
Jean-Jacques Lauble
Posts: 38
Location
Hi Steven

I think that there is no general method to build and analyse models and the only way to discuss positively is with a concrete model and a concrete problem. I think that everything else is a generalisation that is most of the time inappropriate to th problem in question.

JJ
 Subject : Re:Re:Complexity of finding all (major) feedback loops in a model.. 06/04/2019 05:51:20 PM 
Steven Wallis
Posts: 7
Location
JJ,

Amusing. You make a general statement about the limits of generalization.

Perhaps we have different understandings of what it means to have a method for building and analyzing models.

Despite your beliefs, there are a number of such methods for building and analyzing models (e.g. "Systematicity" for evaluating student knowledge, IPA for theories and plans, "integrative complexity" for correspondence).

Of course, they must be applied to a specific model (broadly understood) and that model may also be evaluated regarding its relationship to a specific problem/situation (Ohm's law is very useful for electronics, not so useful for interior design).

Thanks,

Steve
 Subject : Re:Re:Complexity of finding all (major) feedback loops in a model.. 06/05/2019 02:25:47 AM 
Guido W. Reichert
Posts: 24
Location
Steven,

interestingly enough, while Ohm’s law is not helpful for interior design, modeling the choice behavior in selecting furniture can be modeled using Kirchhoff’s laws. So it seems that Forrester’s “home turf” is never far off. ;-)

Some interior design is “rationalized” with the teachings of feng shui which can be seen as making rules about energy flows albeit using some obscure beliefs about the five basic elements...

This talk about generalizations and model abstractions reminds me of Synergetics introduced by Hermann Haken and then used to model social behavior (i.e. sociophysics - Hermann Haken and Wolfgang Weidlich are unfortunately only mentioned in the German Wikipedia article) ...

Personally, I would tend to disagree with the general statement made by Coyle: It may well be leading to something new, if some principles lead to models that are further ahead of our understanding. Then, we have to catch up and learn something from a model we have not understood yet. Alpha Zero plays Go better than any human precisely for that reason. Sometimes always keeping ahead of the model simply is a euphemism for GIGO.

Best regards,
Guido
Last Edited On: 06/05/2019 05:59:50 AM By Guido W. Reichert
 Subject : Re:Re:Complexity of finding all (major) feedback loops in a model.. 06/05/2019 09:45:26 AM 
Steven Wallis
Posts: 7
Location
Guido,

Good points - Especially interesting is the game of continually trying to catch up. Sounds like a co-evolutionary kind of "red queen effect." The race advances relatively slowly for biological evolution; faster for social evolution; and, for concepts, may evolve quite rapidly (as we see in the proliferation of fields/theories/publications/perspectives.

Now, I'll have to explore sociophysics.


Thanks,

Steve
 Subject : Re:Re:Complexity of finding all (major) feedback loops in a model.. 06/06/2019 03:13:11 AM 
Jean-Jacques Lauble
Posts: 38
Location
Hi Steve

To make our conversation useful, it is important that you take into consideration our differences, you and me. I am strictly on the user side, trying to find solutions to my problems, without any preference for any method. If a crystal ball works, I will use it.

I am too on the practical side, being the decision maker and having too to implement the solutions and I know my problems from the inside, having worked practically with it.

I have studied different methods in the past, exposed in books or in distant courses. I have tested them but none have been working. The reason that I have discovered with the time, is that each method has worked in a specific environment, with a specific problem that does not correspond to my situation.
And the environments and problems can be very different.

After having lost years trying to apply methods badly adapted, I realized that I had to find myself how to work with my own problems and resources forgetting the other methods even if I can take some of their ideas.

I will not detail here how I work, but if I wanted to explain it to say a consultant, he would better not use it, because it would not be adapted to the problem he is trying to solve and to the environment he is working in: money and personal time the client is ready to lose, experience of the client, personal interest of the client that mixes with the problem. For example, the interest if the client is a politician will be to win the next election, the interest of a big company C.E.O. will be to produce a good profit the next year, and the owner of a small company will probably be more interested with the long-term effect of the policies. A consultant must take this into consideration even if it does not fit into a strict scientific study of the problem.

This is why I said that there is no general method of working with SD even if the promoters of the method claim the contrary: they must sell their books, find student and clients.

Best regards.

JJ
 Subject : Re:Re:Complexity of finding all (major) feedback loops in a model.. 06/06/2019 03:26:01 AM 
Jean-Jacques Lauble
Posts: 38
Location
Hi Guido

Before understand anything, you will spend some time in a non-understanding phase. I have a different interpretation of Coyle’s advice. Keeping your understanding of a model on a high level does not limit your capacity to improve a model later on. I think that on the contrary a low level of understanding will lead towards bad directions and increased low understanding, not mentioning the time loss.

Best

JJ
 
<< first < Prev 1 2 Next > last >>
# of Topics per Page