Learning and System Dynamics

["Martin F. G. Schaffernicht"]

Hi,

thanks for all the reflection and hints on written stuff. Before diving
into this, I think I owe you some explanations.

I am interested in modeling systems that help people learn, and not in
modeling how people (do not) learn. In this sense, the "learning" I am
looking for does not take into account all learning humans are capable
of.

Certainly learning can be seen at very different levels: the biological
(neuronal) level which for humans is "tacit" is one, and the conscient
level is another. Tacit learning is going on all the time (for example:
habituation), and since the conscious domain emerges from the tacit
level, conscious learning is also affetced by this. Conscious learning
will in general be deliberate. What I am trying to do is use processes
from the "tacit" domain to guide conscious learning.

It seems that when learning is defined as "new response to a stimulus",
this excludes all the times when something gets confirmed; so I prefer
thinking of learning of "internal transformations that change the
conditons for future action". It seems to me that people need something
to recall tham that (inprinciple) each time they decide something, there
is a chance to do so as they did before or to do so in a modified way,
and the results reached by previsous decisions/actions may guide them in
this choice.

So in my mind there are some basic fields of learning that stem from the
basic components of action and learning. In general, action is soething
a system does to relief a tension (a difference between how something is
and how it should be). This takes the "something" to be defined (a
dimension or variable and a target value for it), and some kind of
"action rules" to be defined. Once these components are defined, the
system can act. However, how come that the "dimension" is this one,
that the target value is this one, and that the actino rules are these
ones? In order to obtain some kind of autonomy, the system has to
"know" how to configure the action rules, the target (or threshold)
values and the dimenions (or variables).

Now for people it is usually hard to tell how they configure their
action rules, how they come to fix a target value or a (target)
dimension; however, they (ideally) should be able to tell, in form of
"theory-in-use". So it may be plausible to use mechanims known from
"tacit" processes (control theory?) in roder to guide them towards this
goal (this is a learning in itself).

So I have been trying to formulate target level adjustment with
something similar to the elevator example, but I found that when it
comes to change action rules or variables, it is hard to imagine this in
a system dynamics model. Arrived at his point, I presented my
reflection during the recently held "First Latinamerican System Dynamics
Congress", and thanks to the comments of various of you, I dared to make
the QUERY.

So: thanks again to all of you, and Ill be reading the stuff you
indicated!

"Cordial saludo",

Martin
From: "Martin F. G. Schaffernicht" <martin@utalca.cl>

--
Martin F. G. Schaffernicht
Facultad de Ciencias Empresariales
Universidad de Talca

http://dig.utalca.cl/carpeta/Martin_Schaffernicht

["Strobel, Johannes (UMC-Student)]

Hello,
I very much enjoyed the discussion, fascinating and very diverse ideas.
Just an addition:
Two colleagues and I wrote a chapter Modeling for meaningful learning:
We distinguish in the paper on learning with simulations (using models)
and learning through modeling, the later a much more engaging process.
We also distinguish between different cognitive affordances for modeling
different areas. The paper includes many different forms of computational
models (as found in expert systems, semantic networks, and system modeling
tools).

Jonassen, D.H., Strobel, J., & Gottdenker, J. (in press). Modeling for meaningful
learning. In R. Floden & K. McKevitt (Eds.), Technology for meaningful learning. New
York: Teachers College Press.

Its not available, yet, but if one is interested (s)he can contact me for an electronic copy...
Johannes

From: "Strobel, Johannes (UMC-Student)" <jse09@mizzou.edu>