In the last 15 years the personal computer has been transformed from an exotic luxury into one of the common objects of modern living. Computers, and the services they provide, are becoming an inescapable part of the workplace; and the super-exponential growth experienced by the World Wide Web promises to make "computer literacy" a necessary trait for any member of a "well trained" workforce. Thus, I would seriously argue that an undergraduate student leaving college without significant hands on computer experience has not received an up-to-date (or perhaps even an adequate?) education.
The situation is even more pronounced in the sciences, where computers have become valuable research tools in all fields, and many now speak of a new, interdisciplinary, subject, "computational science", which treats computation itself as a means of scientific discovery. Clearly, anyone aspiring to be a modern research scientist must attain a degree of fluency with computers, and probably a working knowledge of "scientific computation" as well. However, despite the widespread evidence that computational literacy has become a requirement of most scientific research, organized efforts to teach these skills in the classroom have been largely limited to isolated efforts at individual schools and computing centers. If we as a society are to lead the transition to the information age, a coherent, well thought out, multi-institution approach to educating students in these matters is needed.
To address this need, the Department of Energy, through the Office of Scientific Computing, has initiated a set of educational projects in computational science. One of these, the Undergraduate Computational Engineering and Sciences (UCES) project, specifically promotes the incorporation of computation into the undergraduate curriculum. Over forty researchers and educators from the around the country have volunteered their time and effort to further UCES goals. Despite the widely varying backgrounds of UCES' members, a set of common experiences and desires emerged. I would like to present some of these shared experiences, and describe the actions UCES has taken to address them. These include:
1) A "core" set of concepts and techniques which any "computer literate" student should master must be identified. First and foremost, students must be presented with a variety of tools, including symbolic manipulators, procedural programming, visualization software, etc. and encouraged to choose the tool appropriate to the task at hand.
2) Advanced topics for the relative few who will pursue high end research or technical positions can also be addressed, but the emphasis should be on the mainstream.
3) Because computational science is a new field, new educational materials addressing the "core" issues must be developed and made readily available. Given the nature of the subject, it is most constructive if the materials themselves are provided "on-line". This approach also allows easy access by students and faculty at a wide range of institutions, not just research universities and supercomputing centers.
4) Educators themselves need easy-to-use tools authoring tools for creating suitable electronic materials. UCES has made significant progress in this regard.
5) A valid means of professional recognition must be provided so that educators are rewarded for addressing these issues. In order to implement a paradigm shift in how computers are used in education there must be tangible rewards for those who buck the "publish or perish" environment at research universities or take on an extra class at a liberal arts college or teaching school.
6) Other subjects, including our efforts at industrial outreach, as time allows.