accreditation

[Chris White]

I would like to add my support for the accreditation of SD
professionals,
and I especially enjoyed Nelson Repennings latest comment (SD1416).
Granted, some of the details need to be discussed, but the concept is
worth pursuing.

I am a member of the American Society for Quality (ASQ, formerly known
as ASQC, the American Society for Quality Control) and ASQ has several
certified programs that seem to have helped the quality profession by
demonstrating competence (e.g., certified reliability engineer,
certified
quality inspector ). As new areas of focus emerge, new certifications
are
added to address a specific focus. For instance, with the growth of TQM

and the growing role of quality management in general, ASQ added the
Certified Quality Manager to its accreditations in the last 2 years.

As another professional society example, the Project Management
Institute (PMI) has a certification for Project Management Professional
(PMP) to demonstrate a knowledge of the fundamentals of sound project
management. This certification has also seemed to help the field and
give
it credibility.

I see industry benefitting from SD accreditations in the same manner.

Jim Hines asked earlier for any examples of unqualified people sullying
the
reputation of SD. I have one specific example.

I recently observed a presentation by a so-called SD practitioner in
which a
management flight simulator for a specific model was discussed. The
presentation focused mainly on the management flight simulator interface

that was built on top of the model with very little attention given to
the
underlying model structure and logic. All the bells and whistles were
emphasized, and the presenter told how all the customers that bought
this
tool enjoyed its features, etc.

I was finally able to talk with the presenter to get some of the details
of the
model. As it turns out, the model contained very few feedback loops and

instead contained many time-series inputs (based on statistical
correlations) that drove the behavior of the model. For example, one of
the
main outputs was the cost of a project. When I asked how the cost was
determined in the model, the presenter happily referred to an input
time-
series variable called "average cost of project" that started low and
ended
with a large "balloon" at the end of the project. The presenter argued
that
data had shown that most projects cost very little in the beginning and
more at the end, hence, the shape of the input. Amazingly, the same
shape was seen as an output! Sadly, many customers were using this
tool and achieving great "insights".

This presenter was self-taught by using SD modeling software.
Inherently,
theres nothing wrong with that approach. However, there is much more
to
SD than learning how drag-and-drop icons on an interface. But, when
there is money to be made or people to be impressed, there is often not
time to do the real learning.

I am also self-taught using SD modeling software, so my views are not
coming from academia. Yet, to be a benefit to society I want to have a
certain level of competence. Even though my pocket book would be fine,
I
could not be pleased with myself if I were advising others based on the
outputs of models whose validity I doubted. By agreeing to
accreditation, I
run the risk of not being at the proper level of competency, but this is

something I want to know.

J. Chris White
Program Manager
Decision Dynamics, Inc.
4600 East West Hwy, Suite 410
Bethesda, MD 20814
tel: 301-657-8500, ext. 107
fax: 301-657-8626
email: jcwhite@decisiondynamics.com
website: www.decisiondynamics.com

[Jim Hines]

Thanks to J. Chris White for providing a first example of inexperienced
use of SD. The request for examples is important because one of the
arguments for accredation is that accredation will protect the fields
reputation from being sullied by inexperienced people.

Chris does provide an example of a misuse of system dynamics, but NOT of
a misuse that actually ***sullied the reputation*** of the field.

There may be other reasons for accredation, but -- judging from the lack
of examples --- protecting the fields good name does not seem to be one
of them.

Regards,
Jim Hines
From: Jim Hines <jimhines@interserv.com>

[Tom Fiddaman]

Jim,

>Chris does provide an example of a misuse of system dynamics, but NOT of
>a misuse that actually ***sullied the reputation*** of the field.

While the example Chris cited hasnt sullied the reputation of the field
yet, it certainly seems like it has the potential to do so. From Chris
description, it sounds like the model contains misused exogenous
correlational formulations, like the cost balloon at the end of a project,
that have much better alternatives widely available in the SD literature
(e.g. Tarek Abdel-Hamids work on software project dynamics). To me, this
means that the results of the model are at best coincidentally correct, may
be false and misleading, and certainly fall short of the learning that
could come from a better model.

The customer may be happy with the work because they got a lot out of the
modeling process, or because they simply cant evaluate the model quality.
But sooner or later a customer will come along who recognizes poor quality
or suffers a catastrophe due to a misguided strategic decision.

Bad models erode the credibility of all models - and we already have a hard
enough time getting model-based insights implemented. Of course, we deal
with this problem already - managers have an innate (and probably healthy)
suspicion of models in general, learned through experience with thousands
of badly formulated or irrelevant spreadsheets, databases, and other tools.

>There may be other reasons for accredation, but -- judging from the lack
>of examples --- protecting the fields good name does not seem to be one
>of them.

It seems premature to conclude that the lack of examples reflects a lack of
incidents rather than reluctance to share bad examples.

Im actually less concerned about sullying the reputation of the field with
a few catastrophes than I am about experiencing a general low level of
repeat business due to lack of consistently spectacular results. It seems
to me that this is more the mode of demise of most management fads. In this
case, accreditation has to somehow help us get more spectacular results,
and those results have to get attributed to use of SD.

I know several people have modeled the growth of the field (theres an
example in the Vensim documentation, I think). If we think models help,
perhaps we should share and critique some of these efforts.

I think Ive exceeded my ranting quota for the week, so Ill leave it at that.

- Tom

****************************************************
Thomas Fiddaman, Ph.D.
Ventana Systems http://www.vensim.com
34025 Mann Road Tel (360) 793-0903
Sultan, WA 98294 Fax (360) 793-2911
Tom@Vensim.com http://home.earthlink.net/~tomfid/
****************************************************

[Jim Hines]

Tom,

I am still skeptical about the need to protect the fields reputation
from inexperienced pratitioners. A lack of examples may be the result
of a reluctance to share bad news, but a lack of examples is also
certainly consistant with no examples. Putting together a widely
agreeable accredation process and staffing it, is a huge effort. If
accredation does not solve a problem, it is not worth it.

I suspect that business policies are generally so poorly constructed and
so inconsistent that even a novices efforts will lead to some
improvement.

You say that "But sooner or later a customer will come along who ...
suffers a catastrophe due to a misguided strategic decision."
Protecting people from catastrophic SD studies is a much more ambitous
effort than accrediting new folks. I am aware of two SD-related
business catastrophes. Both were caused by SD work done by people who
are very experienced and who would certainly be accredited under any
conceivable accredation process.

Regards,
Jim
From: Jim Hines <jimhines@interserv.com>

[Ralph]

Although I have been an SD modeler for a number of years, I am new to this
listserve. People seem to generate many pros and cons about the
accreditation process. Perhaps I have missed it, but is everyone who argues
one way or the other, generating an argument from their mental models of
the process. The main thing we do as system dynamicists is to develop
models of dynamic processes. This is our strong point. If you are
interested in this issue, the first thing I would suggest is to consider
modeling the dynamics of the accreditation process. What are the
unintended consequences of setting standards at this time in our history?
How will it affect the quality of our models? How will it affect the
existing training programs? Will there be a demand to start an SD training
program at universities and colleges which currently do not have such a
programs? What are the long term impacts on the System Dynamics society?
What would be the impacts on the quality and number of papers submitted to
the System Dynamics review? Would accreditation prevent most catastrophic
SD studies? In general, what are the financial and social costs of
certifying quality.

Currently the System Dynamics Society is formulating a model of the
dynamics of its membership. Would it seem reasonable to enlarge the scope
of the model to include ascertaining the impact of accreditation on
membership, quality of modeling, etc.?

I guess my point is that I get very nervous being around systems people who
at the moment do not seem to be using the tools they were trained to use in
this situation. Also, I sort of felt the same way when so many people began
to talk about transforming firm into a learning organization without
suggesting building a model of the process. A model is something we can
all look at. It would be the means for us to have a greater degree of
insight concerning whether we really want to initiate an accreditation
process at this time.

Ralph Levine, Ph.D.
Department of Psychology
Michigan State University
1-517-332-2317
E-mail: leviner@pilot.msu.edu
Fax:(517)432-2476

[Jorge Nelson]

Ralph:

re:
The main thing we do as system dynamicists is to develop
models of dynamic processes. This is our strong point. If you are
interested in this issue, the first thing I would suggest is to consider
modeling the dynamics of the accreditation process.

Great idea. Let us model what we are discussing.

It reminds me of the International Conference of Technology and Education
(Santa Fe 98) where most of the sessions talked about technology and hardly
anyone used technology at all, besides Powerpoint presentations of what
they will do. I presented an SD simulation and blew away everyone because I
used technology. One gentleman asked me why I wasnt a keynote speaker
because he thought my sim was the only technological breakthrough that he
witnessed during the whole conference.

If we are to be credible we must create the paradigm we profess. I foresee
a model posted on the web where we debate first the model conceptually and
then the outcomes. And, maybe, come to agreement about if we are to go with
accreditation and why?

How can I help?

Jorge Olaf Nelson, Ed.D., Superintendent
American School of Durango, Mexico
PO Box 495, Fco. Sarabia 416 Pte.
Durango, Durango 34000 Mexico
day tel: 011-52-18-133636/115098
eve tel: 011-52-18-134805
FAX tel: 011-52-18-112839
mailto:jonelson@hbcybernetics.com
resume: http://www.hbcybernetics.com
esumes
esume%2098.htm

[Paz Daniel Pablo INOP]

I=B4m totally agree with Ralph.

What happens here with acreditation seems like saying "SD runs for the
outside world not for US"

If the modeller will consider the world from this biased perspective,
the models that generates then will be fatally biased to it preconcepts
and limited vision of the world. We will be generating models of the
world that bring us to conclussions seriously affected by our wrong
visions of the reality as outsiders of it.

Please let=B4s mantain the diversity and tha acess of new ideas !!

Daniel Paz
Industrial Engineering
SIDERAR
EMAIL: apapdp@SIDERAR.COM
----------