June 1993

Albuquerque

Thanks to all who attended the first meeting of the Undergraduate Education Project in Computational Science. It was a very good first meeting and we look forward to seeing all of you again soon.

After each meeting we will provide a short summary of the major issues which were discussed and any actions either taken or decided upon, and will archive these summaries so that others interested in this project can trace its development. These notes will not be "minutes" per se, but rather a sketch of the high points, and the format remains open. Hence, if anyone feels something of importance has been omitted or mischaracterized please contact me and the notes will be suitably modified.

Albuquerque meeting: 5/26-27, 1993

Twelve were in attendance: Chris Johnson, Joe Zachary, Gary Johnson, Dave Zachman, Terje Vold, Dick Allen, Dave Martin, John Ziebarth, Tom Marchioro, Jim Corones, Craig Henriquez, and Ed Oliver.

A lengthy discussion ensued with each person present discussing their reasons for being involved with the project and their opinions on the state of under- graduate computational science. Highlights of this period were:

  1. Chris Johnson and Joe Zachary presented details on how computational science has evolved at the University of Utah. Both a graduate and an undergraduate "minor" have been created which attempt to span the core of the subject. In addition, Joe has created an on-line tutorial class for teaching programming. This class was taught last year and was extremely well received. Joe and Chris are planning a similar approach to a one quarter Introduction to Computational Science class for freshman to be taught sometime in the upcoming year.
  2. Dave Zachman gave a detailed description of the Guided Design paradigm which teaches problem solving abilities by having small teams of students tackle "real" problems with only inter- mittent interaction with the teacher, who acts mainly as a facilitator instead of a source of information. Dave said this approach is very good for teaching independent thinking and problem solving abilities, but is poor at conveying large amounts of information in a short period of time.

    He also described the introductory programming course Pat Burns is teaching at Colorado State.

  3. Dave Martin described the goals and methods of the NEEDS project, an eight school consortium dedicated to making available educational and research materials via the internet. Dave stressed that the information database is broken into "modules" allowing easy downloading of subject- specific information.
  4. Gary Johnson and Jim Corones both stressed that this project is only beginning, that the nature of the project would be allowed to evolve, and that the most important thing was to get as good a "product" out there as possible and then refine it, based on user feedback, to suit the needs of those utilizing the product.

After all of these short presentations had been heard it was widely agreed there is a genuine need for providing undergraduates with skills in the general areas of computational science --- several of the participants said that each year incoming graduate students were progressively less prepared in these areas. A rough outline on how this material should be taught (based, to some extent, on the minor taught at Utah) would be for a general introductory course in programming and scientific computation at the freshmen level, a more advanced course in scientific programming and/or introductory numerical analysis at the sophomore level, and an additional course for juniors or seniors.

It was decided that, as an initial effort, we would take the freshman introductory course Chris and Joe plan to teach this upcoming year at Utah and would adapt it for electronic distribution in a manner similar to the NEEDS project. Chris and Joe plan to organize their course around a symbolic manipulation package (Maple), and the advantages and disadvantages of this approach were discussed. It was agreed that although this would place some limits on who could make use of the package, within the areas being covered most would have access to one symbolic manipulation package or another, so this would not be too onerous. It was agreed that Gary and Chris' tutorials should be ported to Mathematica as well, both to avoid implicitly advocating a specific SM package, and to broaden the range of people who could make use of the materials. One strong advantage of building the class around a SM package is that questions of interface, hardware, etc. are greatly reduced since Mathematica and Maple perform fairly consistently across platforms.

Several other points about this approach were discussed. Noteworthy among these were:

  1. The "course" should be broken up into modules which could be grabbed individually off the internet. This would increase the chances of these materials being used since instructors could use only those modules which were of interest.
  2. A system for responding to user feedback should be put in place as early as possible.
  3. It was felt that the "modules" should also have a vertical structure, with each containing a brief overview of the general subject followed by a variety of case studies taken from a variety of fields.

It was stressed that these introductory materials, if properly broken into modules, could be far broader in scope than any one course would contain, i.e., individual instructors would extract only those subjects suitable for their particular course. Thus, any reasonable subject/materials would merit inclusion in the distribution.

It was also agreed that we must spend additional time trying to define the broad structure of the project. It is important to strive for a common approach that provides enough flexibility to allow for wide adaptation of the materials produced. Also, we will eventually need to reach some agreement on what the "core" material is at each level of the project i.e. Freshman, Soph-Junior etc. We plan to work on these topics actively before the next meeting and make them the major goals of that meeting.

Another major goal is to identify others "in the system", i.e., other teachers, academic researchers, perhaps people from industry, with similar interests. We need all the help we can get on this! In part this means trying to collect other materials that are already available, and may also mean complying with standards others have implemented. People from the other disciplines relating to computational science are especially needed.

We concluded by canvassing all present for their opinions on what an introductory course in computational science should include (in no particular order). This was felt to be a good idea since the course Chris and Joe are organizing, while meant as a "broad based" introduction to computational science, still is rather specific to the interests of the engineering department at Utah. The outline below had been assembled by the time we adjourned.

It was agreed we would try to meet again in July or August, probably in conjunction with a meeting to be held in Denver in late August. It was also agreed we would try to meet sometime soon in Boston, since the food there is so good and many of us have friends to visit.

Graphics/Visualization
Tools
Hardware/Software
Simulation
Modeling
Guided Design
Case Studies
Symbolic Mathematics
Maple/Mathematica/Axiom/Matlab
Programming Techniques
Computational Science
Interdisciplinary
Case Studies
Linear Systems
Overview
Case Studies
NonLinear Systems
Overview
Case Studies
Computing
Programming Languages
Error Analysis
Architectures
High Performance Computing
Networking
Algorithms
Unix
Mathematics
Tools
Spreadsheets
Word Processing

NEAR TERM ACTIONS TO BE TAKEN:

  1. Support for Chris and Joe in putting their course together.
  2. Porting of some of their tutorials to Mathematica.
  3. Creation of an E-mail alias for the group.
  4. An attempt to seek out others who might be interested in this project and, in particular, to bring in people from related fields not represented at this meeting.

    SUPPORT MATERIALS AVAILABLE:

    The following materials are available from the ALBQ meeting:
    1. These notes.
    2. A brief description of Guided Design and a fairly in-depth example taken from a Guided Design class.
    3. Teaching materials from Pat Burns' introduction to programming class.
    4. The syllabus for Chris and Joe's "Introduction to Computational Science" class to be taught this upcoming year (i.e., the course on which our first set of materials will be based).

    NEXT PREVIOUS UP> </A>
<A HREF = HOME CONTENTS