The Curriculum Committee,
PROCEDURE AND PLAN FOR THE WORK OF
THE CURRICULUM COMMITTEE
Introduction
The work in the Curriculum Committee has progressed more slowly than expected although at least we have now started. Based on an e-mail sent to all committee members in April, see enclosure, I have received a number of relatively general responses:
These responses (except for brief comments) are enclosed in full because I think they contain a lot of interesting information / points of view from which we can all benefit. I have itemized and summarized the feedback below. I suggest we decide in the upcoming meeting in Atlanta, on how to proceed.
First the names of the committee members and an indication of the people from which I have received responses. I apologize if I have overseen some response! Most members did not respond with a SDCC-* title and it has thus been a bit difficult to keep track of the discussion:
Curriculum Committee:
Pal Davidsen, Chair
Yaman Barlas
George Richardson
Peter Milling
John Morecroft
Michael Radzicki (responded)
Khalid Saeed (responded)
John Sterman (responded)
Jac Vennix
Jim Hines (responded)
Jay Forrester (responded)
As you can see, I have asked Jim Hines to join us since he is one of the few with experience in IT-based distance learning (suggested by Bob Eberlein who remains on the e-mail list although he is not an official member of the committee). Jim has agreed and I take it that no-one have any objections to that. Moreover, Jay Forrester is on the mailing list and responds very actively. Also, I have asked Prof. J. Michael Spector, Professor of Instructional Design, Development & Evaluation at Suracuse University (http://soeweb.syr.edu/ idde.html) to act as a “consultant”. He knows SD very well, has been working with me for a long time and has a keen interest in commenting on our work. His introductory note, quoted in full below, should be an indication of his commitment and the interest that he takes in our work. He will, consequently, also appear on our list.
From Michael Spector, I have received references to a number of web-pages that I think might be very useful to us in our further work. They may well serve as guidelines:
1. ACM site on considerations for a new curriculum for computer science (looking at how
another group addressed a similarly complex task might be insightful):
http://www.acm.org/education/curr91/
2. A WebQuest resource site on curriculum design (aimed at online learning but still somwehat informative):
http://www.biopoint.com/WebQuests/
3. An instructional design tutorial at Duke (called curriculum design):
http://www.aas.duke.edu/teach/consult/curdes.shtml
4. Other tools at the University of Washington:
http://www.csd.uwa.edu.au/HERDSA/abstract/curriculum.html
5. Curriculum design tools from Harvard:
http://learnweb.harvard.edu/alps/tfu/design.cfm
Let me tell you that I have a feeling that we are only at the very beginning of a very, very long journey. The responses that I have had to the first report indicated so, and I believe my task is to communicate and systematize the various points of views that have been raised to evaluate whether there are good reasons to move on, - and find ways to do so. I believe we all need to break down our effort into tasks that can be handled within the time and resources we have available. Based on the upcoming discussion that we will have in Atlanta, we should outline a way to move forward.
Summary of responses:
Comments in general
All comments have been generally favorable and it has been stressed that the work of the committee is important. A number of warning flags have been raised though and there is reason to consider them seriously.
Comments on our work in general and on how general our approach should be
John Sterman (1) comments that it is a major challenge to negotiate between a curriculum on the System Dynamics method as a general approach to scientific investigation and a curriculum that is well-founded in a domain of application with which the students will be familiar. In fact he states:
“If this is correct, it might suggest that the project of determining a standard curriculum is at best much harder even than Pål describes, that it might be impossible or ignored by those we might wish to use it, or, at worst, could be harmful by encouraging a standardization that makes the material less relevant and interesting to students.”
Chairman’s comment:
This is, as I see it, a very crucial point. I totally agree that the students can only learn System Dynamics in the context of an application domain with which they are familiar. And, unless there are serious objections, I assume we all do.
When assuming, in my first report, that “ … we are developing a curriculum in system dynamics in general , - not in a domain of application… . it was to try to focus / organize the work of the committee, - a work that can easily expand into infinity.
I would suggest that the committee, at least for now, should agree on a framework of method-specific topics that the curriculum eventually should encompass, - in theory as well as in the form of applications. Although examples / applications are not called for at this stage, they are more than welcome.
If we agree that a methodological topic, say validation, is important, and a particular technique, such as extreme-condition testing, consequently should be included in that framework; then we may;
- illustrate the use of such a technique using examples from various application domains with which the student is familiar;
- discuss and illustrate various ways of teaching or facilitating the learning of extreme condition testing;
- present the theory behind extreme condition testing;
- etc.
I do not believe there is a conflict between creating a method-oriented framework on the one hand and fill that framework with illustrations from various fields, on the other. In fact John’s book, I think, is a great illustration of how to do so. I suggest we proceed that way.
Jack Pugh (2) comments on this topic as well:
“While John's observation about different audiences being better served by different examples, I don't see this as a show stopper. A general curriculum could suggest a type of model or several particular models addressing different audiences. I would not expect any instructor to follow our recommended curriculum precisely, but rather view it as an example and challenge.”
Moreover, Mike Radzicki (2) indicated that this issue is something we all struggle with:
“For whatever it's worth, we here at WPI wrestled with the very question that this committee has been established to answer when we created our B.S. degree in system dynamics. That is, we asked and answered the question: What should a properly trained system dynamicist know? I won't claim that we arrived at THE answer, but only that we arrived at AN answer.”
Some Remarks on Curriculum Development by J. Michael Spector
As indicated above, I have consulted Prof. J. Mike Spector, Syracuse University, on curriculum development in general, and he has offered some very good advise that I would like to share with you:
“You probably already know that curriculum design is a complex activity. First, the word 'curriculum' is understood differently in various places and circumstances. It sometimes refers to a related collection of readings. It might refer to a collection of courses - also called a learning plan or program of study in some places. It might refer more broadly to readings, courses, and learning experiences and activities, including the various instructional approaches involved. The middle use is perhaps the most common, but it is easy to slip into the more narrow or more broad use. This ambiguity can cause problems at the outset in curriculum planning.
One way to help clarify things at the outset is to ask questions about what sorts of competence and expertise are involved, who requires such competence and expertise, what sorts of situations involve such skills and understanding, and what sorts of people might be interested in acquiring those skills and understanding.
It is important not to slip into lesson planning and course design too early. At the course and lesson level, it is appropriate to address issues concerning types of learning goals and appropriate instructional strategies. At the curriculum planning level, it is important to keep a focus on the overall set of skills and knowledge involved.
For example, if we think about pilot training, we can address the competence and expertise involved (take off safely, fly in various weather circumstances, land, and so on, all with regard to a particular airplane). Pilots and co-pilots requires this competence. Flying situations vary involving different airports, varying weather, and so on. People interested are between the ages of 20 and 50, typically have college degrees, and so on.
It is quite common to then conduct a task analysis to elaborate the knowledge and skills involved. Next, it is important to realize that people simply do not learn things all at once - especially not complex things. So there should be some analysis of how the various skills and knowledge are interrelated - again still at a fairly high level and not at the discrete level of learning what a stock is, for example.
I highly recommend Jeroen van Merriënboer's Training Complex Cognitive Skills to help understand this process. While van Merriënboer is an advocate of whole-task training and holistic understanding, he is quite sensitive to the issues involved in task analysis and the need for identifying skill clusters and associated supportive knowledge. He distinguishes between recurrent and non-recurrent skills. Recurrent skills are performed in roughly the same way regardless of changes in the situation. Non-recurrent skills are performed differently depending on critical aspects of the situation in which the task is performed. Most complex skills involve both types. Instruction to support the two types is quite different. For example, instruction for recurrent skills might focus on the acquisition of procedural knowledge and automatic performance of the associated skill in conditions where there are distracters. Instruction for non-recurrent skills typically focuses on heuristics and provides a great deal of variety in selecting problem situations in which to engage learners.
Much learning in and about complex systems can be situated around authentic problems. But the method (case-based learning, problem-based learning, situated learning, etc.) should not drive the curriculum. The curriculum should drive the methods and approaches used. Most likely, multiple methods and approaches will be relevant.
If one attempts to do too much in a curriculum design effort, the effort is likely to fail. I am thinking that curriculum design does not occur in a vacuum. Other curricula typically exist and have been created according to various curriculum models. Students might take courses in more than one curriculum and have already developed some expectations with regard to curricula and courses. In short, some recognition of the surroundings in which the curriculum exists is essential. One cannot generally remake the world at large.
I doubt that this means that anything significant must be given up in system dynamics education. The focus on entire systems should be retained and introduced at the very beginning, most likely. Most likely, it will be effective to begin with a situated example of a simpler complex system that is likely to be accessible to many different learners (a small environmental model, for example). It is important to help those new to a domain to chunk ideas and develop skills.
With regard to Bloom's taxonomy, you already know that it is aimed primarily at instructional objectives - typically at the module or lesson level - rather than at professional development, although the six stages are often presented at both levels. What the Dreyfus brothers have presented is also relevant in this regard: novice, advanced beginner, competent performer, proficient performer, expert. These are more clearly developmental stages in a skill acquisition model. I also think that Tom Shuell's 12 learning activities/functions are worth considering when one gets past the curriculum analysis and into instructional development:
· Expecting
· Becoming motivated (or not)
· Activating prior knowledge
· Attending to
· Encoding
· Comparing
· Generating hypotheses
· Repeating
· Receiving and sending feedback
· Evaluating
· Monitoring
· Combining, integrating, synthesizing
And, at the course and lesson development stage, the 1990 paper by Gagné and Merrill with the notion that most learning is aimed at an integrated and purposeful collection of related human activities (they called that an enterprise). Examples might include replacing an aircraft engine or finding the equilibrium point in a complex model. The first of these probably involves mostly recurrent skills whereas the second might involve more non-recurrent skills (a guess on my part without having conducted a thorough or thoughtful task analysis).
If system dynamics is to grow outside the current small but elite community, then it will be increasingly important to recognize that not every potential system dynamics student is as smart as John Sterman or Jay Forrester. Nor is it likely that every system dynamics student needs to "fall in love" with system dynamics.
Keeping a focus on competencies - skills, knowledge and relevant attitudes - at the various levels at which system dynamics curricula might be implemented will keep the effort on track. Moreover, it is important to keep an open mind and to plan an iterative curriculum development process that includes formative evaluations along the way. Whether one begins with dynamic behavior or with causal representations of whole systems might not make as much difference overall as whether they are both covered somewhere along the way, for example.
Last, get all of the goals and assumptions of the curriculum design effort on the table as early as possible. Many will be implicit. I notice that you have already made an effort to make some expicit. I would guess that many goals and assumptions remain hidden. I have already alluded to one goal that I expect has yet to be stated in an explicit manner - encouraging the growth and spread of system dynamics in high schools, colleges and universities. Once that goal is stated, then it will be manifestly clear that one should not aim all curricula for MIT students.
Well, those are my first thoughts. Please feel free to share them with your curriculum committee.
Michael Spector “
We need to set clear goals for our education as a whole as well as for the various components
that altogether constitute the SD curriculum.
Jay Forrester (5b) is pointing out the need for us to identify more clearly the purpose of the education, in particular in terms of a target student population and the length of the study:
“ So far, the communications about a system dynamics education imply a single educational sequence. However, I believe that for several years, and probably forever, there will be widely different scopes and purposes for a SD education. The outline that has been started
by Davidsen suggests an extended and deep education toward becoming a system dynamics professional.
I do see system dynamics as a profession that can be at least as demanding as engineering or medicine. However, all the professions have education for different purposes and different levels of expertise. We should be thinking along similar lines. For different purposes, there is still the question of what aspects of system dynamics should be included, and what is the expected outcome for the different programs.
As examples:
1. System dynamics as part of a one-year or a two-year program focused on some other professional objective. The present teaching of SD as part of a master of business education falls into the category in which SD is only part of the objective.
2. A four-year undergraduate program leading to a B.Sc. degree in system dynamics. We now have an example at Worcester Polytechnic Institute.
3. A four-year (or more) Ph.D. program in system dynamics for students without prior system dynamics education. Is such a doctoral program importantly different from the four-year undergraduate program? Will it lead to a level of expertise different from the "intermediate level" described by Radzicki?
4. A four-year Ph.D. program for students who already have four years of undergraduate SD training. As far as I know, this would be material beyond anything that is now being taught.
5. A 12-year SD program threaded through a K-12 pre-college education. Most of this has yet to be developed.
6. A four-year undergraduate program for students who have had a good foundation in SD during their K-12 years. Such would be new material.
7. A four-year Ph.D. Program that would accept students who have had 16 years of SD exposure from kindergarten through undergraduate college. Such would be far beyond anything that is now being developed.
8. And then there is SD as a cultural subject and a philosophy and way to look at life and the world. Such could fit anywhere, but should be included in a liberal arts education. For some of the objectives, see my attached paper, "Learning through System Dynamics
as Preparation for the 21st Century."
To be sure, the more advanced programs above lie rather far in the future, but it is not too early to speculate about what they would be like. It is only by establishing the vision of where we are going that we can encourage people to begin the steps along the way. ””
Mike Radzicki (3) is also concrete on this:
“The goal of all of this [the WPI curriculum] is to train someone who can produce an original SD model for a public or private sector organization, either individually or in a group setting, and get the organization to adopt the policy recommendations suggested by the model.”
I might add that, despite all of the training described above (and with the caveat that we haven’t had THAT many students go through the program yet and that there are exceptions), we have learned that the students are at best “intermediate level” (as opposed to “expert level”) modelers when they graduate. More specifically, by the time a student is a senior he/she knows enough to avoid rookie mistakes (e.g., no first order control, wrong variable dimensions, etc.), but usually cannot build a model ON THEIR OWN that yields insights. In other words, they build models that are not technically incorrect, but that don’t yield any new information. My personal view is that this is because they either make their models too
aggregate or because they do not build in enough policy space.”
Khalid Saeed (4) reinforces this argument:
“The second point concerns integration of practice with a problem context, which we have attempted at WPI through requiring an application area. Text materials that strongly link practice with an application area are indeed important and I particularly commend John Sterman and Andy Ford for their recent attempts to achieve this union. There also exists pioneering work in Mass: Economic Dynamics, Lyneis: Corporate Planning and Policy Design, Forrester: Urban Dynamics and Industrial Dynamics in that tradition. I expect SD will be applied extensively in learning social sciences in future. More text materials focused on specific social sciences would be very helpful to facilitate this.”
Chairman’s comment:
I think Radzicki offers a great example of a goal that applies to the entire SD education. The goal set by WPI cannot be reached unless the “pure” SD education is embedded (actually based on) a number of skills already acquired by the students and unless the students are allowed to work in the context of a public or private enterprise, - as is the case in the WPI program.
This indicates how the goal sets, not only the content, but also the context for and, most likely the method to be followed in, the SD education.
Mike Radzicki (3) offers important experience:
“A further lesson from our experience at WPI is that there is NO SUBSTITUTE for mentoring. Without an experienced faculty member working with a student on his/her SD project, the results will usually be terrible. BTW: If the Society wishes to promote the creation of better SD models/projects around the world, its senior members need to stand ready to informally look over models produced by rookies (and sent to them over the internet) and suggest helpful changes.”
Khalid Saeed (4) also reinforces this argument:
“The first is learning through personal apprenticeship of experts. There is indeed no substitute for it. When a substitute is found, the practice would become so regimented that it would not create new insights. Teaching system dynamics will remain labor intensive if it is done well.”
Chairman’s comment:
Personally I fully agree in principle. In times of internet, it becomes an challenge to find modes of interaction between teachers and students that effectively facilitates asynchronous mentoring at a distance. Moreover, it seems obvious that there is still much to be gained from software supported facilitation during the modeling process.
John Sterman (1) more specifically addresses teacher vs. learner directed learning, here in the context of generalization:
“Arguing up front that the tools and method are general and that what you learn about some particular example will apply to many others is not the way to go. Students should be presented with examples and cases that are already close to topics they are interested in, and then discover for themselves (helped by the teacher) that these ideas apply far more broadly. The ideal situation is for the student to tell the teacher that "you know, your model is much more general than you think" - when that happens, it is glorious - That is learner-directed learning!”
Chairman’s comment:
Again, I fully agree with John in his learner-directed learning approach calling for the students to discover various aspects of the SD approach in general and SD models in particular. So the question is then, what are they to discover? What kind of insight do we expect the students to gain from the various examples presented to them? If we can name them, - then that would be a part of the framework that we, as I see it, are about to develop.
Finally, the “act of generalization”
is cognitively pretty sophisticated, yet one of the many cognitive
skills we want our students to develop, - whatever we mean by
“dynamic intuition” being another one. I think that the
development of such cognitive skills should be learning goals in our
curriculum. Having stated goals of this kind, we need methods by which
we can to reach those goals. In a teacher-directed curriculum, these
methods are those of instruction. In a learner-directed curriculum,
they belong to the method to be learned, i.e. in this case to System
Dynamics itself. So, generalization, say, is an element in the system
dynamics method and part of what should be learned.
Comments on spiral learning and content
Learning System Dynamics is a process that needs to be fueled by theory, practice, contemplation, reflection etc. One way of envisioning such a process is by way of a spiral (i.e. spiral learning), - as suggested in the first report.
Jay Forrester (5c) offers some very important views on spiral learning among them on the spiraling velocity. In that context, he suggests some topics to be covered by a spiral-based curriculum:
“In Davidsen's papers to the committee he properly calls for a spiral approach to system dynamics education whereby a topic will be revisited many times. I agree that the fundamentals of system dynamics will not be learned and appreciated in one exposure. There should be several passes with increasing depth and completeness.
Pal refers to the spiral approach in the context of the Bloom taxonomy. That is one kind of spiral. A single turn of the spiral could take place with each and every model beginning with the simplest first-order models and moving through oscillation and into more complex concepts. Such might be called the high-speed spiral because it would occur so often and each turn might be completed in a time span from a day to a month.
But there is another kind of spiral that would keep passing through the fundamental principals, concepts, and good-modeling practices of system dynamics. A turn in this longer spiral might occupy a month or more.
We adopted this latter spiral idea as a framework for the Road Maps series of papers that are on our web site, sysdyn.mit.edu. We may not have succeeded as well as I had hoped in following the spiral structure, but the ideal was much discussed and we tried to implement it.
In our attempt, we visualized the spiral as a structure of vertical rods standing in a circle with a spiral helix climbing around the outside of the rods. Each rod represented something to which the student should be exposed. Not every rod would be treated in every turn of the spiral. Our choice of topics for the vertical rods were of many different types of ideas, principles, and practices. The following is illustrative, but far from a complete list:
Modeling exercises, building models
Mental simulation
Generic structures
Numerical, hand computation of simple simulations
Nature of complex systems
Mistakes and misunderstandings observed in models
Tests for building confidence in models
The economics of model building and testing
Oscillation
Steps in model building
Sensitivity analysis
System dynamics in social policy
System dynamics as a new basis for teaching economics
Fundamental principles (many topics along the lines of Forrester, J. W. (1968). Principles of Systems. Waltham, MA, Pegasus Communications. Available from Pegasus Communications, Inc., One Moody Street, Waltham, MA 02453-5339, tel: 781-398-9700, fax: 781-894-7026
There will be more to be added, but the idea is to identify the concepts that should become second nature in the thinking of an accomplished system dynamicist.””
Chairman’s comment:
In offering Bloom’s approach as an example it was because I think he has a lot to offer in terms of progress in cognitive processing. I leave this for the paragraph on Bloom’s taxonomy.
I think it is possible to define a slow-turning spiral and a number of fast-turning spirals within each, - typically addressing a particular technique, say, from practice to theory and back to practice.
The topics listed by Forrester are clearly concrete examples of what needs to be covered. The question is whether we may organize such a mixed bag of topics in such a way that they lend themselves to a systematic curriculum construction.
We need to progress in learning with respect to a variety of knowledge domains and skills, - to mention a few;
- behavior patterns;
- structural size and diversity (complexity) (variety of relationships);
- knowledge elicitation, research and ‘listening’-skills;
- modeling skills;
- integration skills and simulation supported (from pen & paper to computer) integration skills (dynamic intuition);
- confidence building, validation and sensitivity testing skills;
- trap avoidance and documentation skills;
- generalization and application skills;
- communication and presentation skills.
Comments on Bloom’s taxonomy
There are many ways to progress in learning. Blooms taxonomy has been offered as an example of how to classify advancement in cognitive skills. As indicated by Michael Spector (see paragraph Some Remarks on Curriculum Development by Michael Spector) there are a number of other ways to classify cognitive skills.
Forrester (5d) objects to the use of Blooms taxonomy in the sense that he believes the sequence in which learning should take place, using system dynamics, should be reversed:
“Davidsen makes the interesting proposal that we follow Bloom's sequence of teaching, starting with knowledge, and later reaching analysis and synthesis.That approach to education may have been necessary in an earlier day when there was no effective way of teaching synthesis. But now that a system dynamics model is an exercise in synthesis, an alternative education becomes possible.
In her 1978 paper, Nancy Roberts discusses departing from Bloom and instead adopting the ideas of learning facts through structure as argued by Bruner: Roberts, N. (1978). "Teaching Dynamic Feedback Systems Thinking: an Elementary View." Management Science 24(8): 836-43. Bruner, J. S. (1963). The Process of Education. New York, Vintage Books. In her paper, Nancy Roberts makes the following observations:
"The area of social studies in the elementary school curriculum has been criticized for years as a discipline whose focus is primarily on facts. Ten and eleven year old children have traditionally been presented with a series of history and geography facts that are quickly forgotten. Bruner suggests these phenomena might be due to the absence in present curricula of any attempt to teach the basic structure of the subject. This pilot study attempted to develop a teaching method that takes into account Bruner's suggestions of focusing on the underlying structure of a subject matter. The fourth and fifth levels of the taxonomy are congruent with dynamic feedback systems thinking. Bloom suggests that to raise a person's level of thinking, that person must go step by step up the hierarchical structure of the taxonomy. This would mean starting at the first level, Knowledge (fact accumulation). Most elementary school teaching materials, especially in the area of social studies, concentrate at the lower end of the hierarchy of thinking processes, with higher levels of thinking in students not resulting."
I think that much of the criticism of education grows out of teaching facts before there is any evident reason for knowing the facts. Such an education is uninspiring and lacks motivation. System dynamics makes possible a structure of education that starts at the analysis and synthesis levels and backs into the lower levels when the need for facts becomes apparent.
I suggest that we reverse the Bloom taxonomy by making the last three of his stages first:
analysis
synthesis
evaluation
and let these lead to the first three of Bloom's stages:
knowledge
comprehension
application.”
Chairman’s comment:
I think there is a common agreement among system dynamicists that synthesis is at the heart of System Dynamics. Consequently, I think we might even put synthesis at the top as a motivator and driving force in SD-based education. It certainly is a driving force behind the quest for knowledge and comprehension. So, in terms of learning goals and priority among them, we are in full agreement. Also, in terms of method, i.e. allowing synthesis to be the overall mode of learning is indeed the way to go, - in my opinion.
The question is just whether, at the micro level, whatever we synthesize (say, model components) must be recognized (known) and comprehended (understood with respect to structure/ behavior relationships) before a synthesis can take place. This is a topic for a very interesting and important discussion that we need to address as well.
Let me just briefly close by referring to an interesting web-page beyond Bloom's taxonomy (Scardamalia and Bereiter): http://csile.oise.utoronto.ca/abstracts/Piaget.html
Comments on context
To understand why a “pure” SD curriculum has been designed the way it is, it is interesting to know the context in which SD is embedded. I.e. what are the students expected to know before they read SD and in which context their are expected to apply their SD experience.
Mike (3) is concrete on this:
“Broadly speaking, system dynamics majors at WPI take courses in the following areas:
1. Mathematics (calculus through ODEs, control theory, and stats/econometrics)
2. Programming (C/Java)
3. Management (accouting, finance, OB)
4. System Dynamics (four "pure" SD courses)
5. Psychology [Specifically tailored to SD] (Group Moidel Building and Judgement and Decision Making)
6. Other Social Science (economics, political science, public policy)
7. An "Application Area" (which can be ANYTHING) (e.g., a student who wishes to get into the production of SD software may take an additional set of CS courses; a student who wished to get more into business administration might take an additional set of MGT courses, a student who wished to get more into psychology might take an additional set of psych courses, etc.).
In addition, WPI students must complete three major thesis-like projects (one in the sophomore year, one in the junior year, and one in the senior year) prior to graduation. The sophomore year project must address some topic in the humanities. The junior year project must focus on some interaction between technology and society. The senior year project must be an application of SD to the student's "Application Area." Thus, a SD student must do at least one major SD project (senior year) and can elect to do two of them (i.e., do the junior year project via SD). [Actually, if a student can find a humanities professor with an open mind, he/she can do a third SD project (we've had one kid do this so far)].”
Jack Pugh comments briefly on this:
(a) one a prerequisite for WPI's numerical analysis course
(b) java is helpful to know when producing web-based simulation environments
(c) we wanted our students to have a bit of an understanding of what lies underneath the various user-friendly SD packages that exist today.
That said, I agree with Jack that programming skills and SD modeling skills are different animals.”
I think it is important to keep the educational context in mind. Our task, however, as I see it, is to concentrate on the “pure” SD discipline. The question is whether we are able to keep this effort going “out of context”. In a sense, therefore, can one talk about a SD discipline and thus a curriculum regardless of context. It is my opinion that one can and that the context offers a number of ways in which SD manifests itself. So it is my suggestion that we identify topics that a SD curriculum should cover (at some appropriate (to the student and the context) level of sophistication) regardless….