Distance Learning for Professional Programmers:
Distance Learning at Carnegie Mellon University
Software Careers Fall 1997 Dr. Dobb's Journal
by James Tomayko
The economic upturn we are happily experiencing comes at the end of a period of severe pain and suffering across the industry. But the shortage of employees means that there are fewer opportunities for full-time supported graduate study, and part-time students trying to attend traditional course offerings are trapped by geography. Realizing this, the School of Computer Science at Carnegie Mellon University, like many other universities, is bringing the courses to the students-no matter where they are or when they need to study.
Interestingly, Andrew Carnegie founded the original Carnegie Institute of Technology at the turn of this century to give the worker a means to improve himself. For over half of the school's history, night students outnumbered day students; degrees were earned through part-time study. Through distance education, Carnegie Mellon University is returning to its roots. Peter Drucker said recently that the big universities will be dead in less than half a century. Technology makes it possible to teach even the most complex concepts to students based anywhere in the world. The universities will feel the backlash from over two decades of well-above-inflation cost-increases.
The first arena in which the new virtual universities will excel is in delivering graduate education to practitioners. These are often the most motivated students, and the ones most in need of just-in-time learning. For all their faults, universities are still centers of leading-edge research and contain critical masses of intellectual power. They will be the best deliverers of graduate courses. Even more-established distance education organizations like the Open University maintain a full-time faculty and a research base in Maynard Keynes, a place in Great Britain few of their students have ever seen.
Carnegie Mellon plans at least four tracks leading to a certificate-software engineering, human-computer interaction, language technology (machine translation and speech recognition), and machine learning. The more-established programs, such as software engineering, are researching ways to offer a complete professional masters degree through distance education that will be indistinguishable in quality from the on-campus program. It can still be argued that the full-time students in residence will gain more from casual interaction with classmates and faculty than any distance education can provide. However, if spending a year on campus is impractical, then remote learning is the solution.
The technologies that make distance education programs viable are largely software technologies. Some distance education courses rely on only e-mail as the medium of exchange, while certain point-to-point courses, such as Carnegie Mellon's program in computational finance are taught from special classrooms connected via T-1 lines. The School of Computer Science decided to adopt the philosophy of using the lowest common denominator technology except in cases where special software is provided by the school accompanying the media. This is intended to make the courses available to the largest audience without sacrificing quality. It enables students from literally around the world to participate.
The base technology is the World Wide Web. In addition to simple home pages using basic HTML, special software is in place for student lounges and work areas, as well as means for the automated collection and display of data, such as the Personal Software Process (http://www.psp.itc.cmu.edu/). Most of this special software was built using Java or Perl and does not represent any breakthroughs.
Perhaps the most impressive technology is Just-in-Time Lectures (JITL). The original idea for this teaching system was by Raj Reddy, dean of the School of Computer Science. It has been refined over the years by several researchers, most recently Roger Dannenberg. There is now a consortium devoted to improving the technology that has industrial members who license its use. Instructors are videotaped, and their slides are gathered in Powerpoint or other presentation media. The slides are juxtaposed with the instructor into one window of the JITL display using advanced video compression techniques. QuickTime for either the Macintosh or Windows is used as the playback software. The presentation is synchronized with a window containing the lecture outline. A third window is a view into the FAQ file. The early ratio of CD-ROM production-to-lecture time was in the neighborhood of 60 hours to one hour. This has been reduced to about 20 to 1, making the technology usable in a rapid production environment. CD-ROMs contain the player software, Netscape, QuickTime, Eudora, and the installation executables, making them highly portable.
The chat software is built and maintained by SneakerLabs, a startup begun by a Carnegie Mellon Master of Software Engineering graduate. A Java applet downloads to the user's machine and enables conversation by text. Again, no special software is obtained by the student: It is supplied on the web site. SneakerChat can also be used by students to form study groups outside of scheduled chat hours.
There is, of course, a growing amount of chat, voice, video, and other types of conferencing software for the Net. SneakerChat is simple, so it fits the philosophy of lowest-common-denominator technology, and is an effective teaching tool.
You can see that even these relatively simple but ubiquitous technologies have the power to transform the business of higher education, particularly continuing professional education. Drucker and Gene Rochlin, author of Trapped in the Net (Princeton University Press, 1997), see both the possibilities and the dangers. Rochlin's model of the "computer trap" has four parts-the lure, snare, costs, and long-term consequences.
The lure of computing technology for distance education is that even the simplest models provide greater interaction and content sophistication than "correspondence courses" that universities have had for decades. The snare is the availability of useful software that runs on vanilla hardware world wide. The cost is the growing dependence on the software technology for delivery of courses. The long-term consequences can only be guessed at, though critics of the technology and faculty Luddites worry about the lack of face-to-face interaction and the possible loss of teaching jobs as a few "star" professors corner the market in popular courses.
What will probably happen will be a mirror of the shakeout in the computing industry in general. At first, everyone will enter the distance education market. Eventually, a few universities will emerge as the dominant players, and these universities will probably be ones that no one today would be able to identify. Other universities will control niche markets, such as particular majors, and their distance education programs will survive on these successes. Finally, most of the early entrants will die out. No one knows how resident education programs will be affected by schools like Harvard offering degree programs online. Residential undergraduate degree programs will probably survive as long as their educational purpose is closely tied to their social purposes.
It is, truly, a time of rapid change in education. The impact of computing technology in distance education is going to be much greater and more effective than the PLATO-like CAI software of a couple decades ago. Those early adopters among the universities who make a commitment to develop distance education will have a major influence in the next century.
DDJ
