Sign In

Communications of the ACM

Log on Education

Revolutionizing the Traditional Classroom Course

When the dot-com revolution was in its full funding era, my company, Cognitive Arts, made a deal with Columbia University to build Columbia online. This was no ordinary "lets hurry and get our university online" kind of deal. I am an educational revolutionary and Columbia knows that. Columbia wasn't in it for the quick buck. It wanted to offer high-quality courses to the world outside its campus. Columbia wanted high quality and so did I.

What did our work have to do with revolution? Well, how good are university lecture courses anyway? Do either the students or the faculty believe lectures represent the best way to educate inquiring minds? Or are they more a matter of university economics, putting many fannies in the seats, without a hint of learning by doing or the requirement to think really hard about something?

The Cognitive Arts/Columbia deal was intended to change all that, and, indeed, the courses produced (and still being produced) are rather extraordinary. How many will be produced is difficult to estimate, since in the mind of those savvy venture capitalists, e-learning went from hot to not in a matter of months.

Although the end result was mixed—great courses and endless budget cutbacks that prevented them from being all we had hoped for or as many as we had hoped for—the lessons we learned from the experience are powerful and important. The primary lesson: whether or not the computer is a revolutionary device that can change education as we know it. There aren't many folks in the university world who believe there's a necessity for a revolution. Universities are not in trouble as they stand. Why then the push to offer courses online?

Before a university decides to create an online presence, it ought to understand exactly why this is needed. Typically, university administrations care about revenue, prestige and image, and not being left behind. Columbia was unlike other universities I encountered, wanting to be part of a great educational experience, wanting to lead rather than follow. The faculty was another story. The faculty of any university is a diverse set of people with diverse sets of interests. And as we all know, while professors care about teaching, their highest priorities are their own research and the approbation of their colleagues. So, our first issue was getting faculty to care about working with us. One cannot simply assume that a given professor will care enough about teaching to be willing to put in the necessary hours to create an online course.

Scholars from Aristotle, to Galileo, to Dewey, to Einstein have pointed out that real learning is by doing. The computer is a learning-by-doing device.

Second, simply deciding to offer an online course doesn't really tell you what gets incorporated online. Existing university lecture courses, the most likely candidates for being offered online since they are most highly subscribed, and in general not that valuable educationally, cannot be incorporated online except in the most uninteresting ways. The usual idea of putting lecture notes and assignments on the Web and having a teaching assistant available to answer questions strikes me as being not much better than a correspondence course. If the idea behind an online course is to provide a high-quality educational experience then we have to ask if what we are copying is a high-quality educational experience in the first place.

Curiously, many computer science departments that ought to know better have been offering courses online by videotaping the lectures and grading the assignments online. One is inclined to ask: What is the role of the lectures? Do lectures contain material not found in the textbook, for example? Many a university student has understood that he or she can skip the lecture and still do well when the assignments are independent of the lecture. So, while one might attend a well-rehearsed lecture on art history to see the pictures and to be entertained, a lecture on C++ is unlikely to be of more use than the text, unless of course, the teacher answers questions the text fails to explicate.

A good place to start then are courses that are doing-centered in the first place. Since learning by doing is an educational philosophy nearly impossible to implement in the traditional classroom, and since the computer is a doing device and a device that can run some pretty fancy simulations, the problem is either choosing doing-based courses or converting courses that have no doing into ones that do. But, no matter how clever a course one builds, and how doing-centered it is, one needs to understand the market for the courses one builds. The realities of this marketplace are rather concerning.

A third issue in offering online courses is attempting to understand why anyone would want to take a given course in the first place. The answer may not be what you expect. Two of the first courses we incorporated online were a Java and a C++ course. These were designed to be learn-by-doing courses that had no lectures and instead relied upon realistic situations in a simulated workplace and interactions with an online mentor who evaluated successive approximations submitted by the student until he or she was satisfied the job was well done. At that point, the next layer of the job unfolded before the student. Context was everything, and help in programming was readily available. We considered this to be a high-quality course (many students loved the course; others thought it was too difficult) and began to sell it to high schools as a way of enabling them to provide a kind of course they were unlikely to teach.

One of the first high schools we sold it to complained bitterly when the question of college credit was asked. Since these courses are in the school of Continuing Education at Columbia University, they carry Continuing Education credits, which are usual for university outreach programs, and is not a Columbia University credit (which shouldn't matter since such credits are only useful at Columbia University). When the high school discovered this it withdrew its students. Why, I asked, weren't students learning to program? They were, I was told. That wasn't why they were taking the course. They were taking the course to get a college credit more cheaply than they would by actually going to Columbia. Students saw this as a way of saving money. Expanding educational opportunity was not their priority.

Accreditation is a big issue. It isn't courses students want online, it's certification.

Consider chemistry. I thought it would be nice to offer an online college chemistry course, but then found myself in the following conversation:

  • Me: How many of your first-year chemistry students are pre-meds?
  • Professor: All.
  • Me: How much of what they learn in first-year chemistry is useful in the life of a doctor?
  • Professor: None.
  • Me: Is there chemistry a doctor needs to know?
  • Professor: Yes.
  • Me: Then why can't we teach that?
  • Professor: Maybe we could, but the pre-med certification process might not say it was acceptable and then students wouldn't take the course.

Ah, naive little me.

How about physics? A physics professor contacted me. He was worried about basically the same issue. There was a course in physics especially for pre-med students and he was shocked to see it didn't contain the basics about physics that a doctor would need to know. He wanted to build something better. We realized we couldn't since any new material we incorporated wouldn't be relevant to the pre-med exam, and after all, pre-med students are interested in getting into medical school, not in learning physics. We did design an online physics course, but it wasn't the one either of us had in mind at the outset.

How about another subject? Economics is a big hit in college these days. Why? Because students want to go into business. It seems they all want to open their own dot-coms (or did). Would the economics professors be interested in teaching an online business course.

How about psychology, the second most popular major? I had noticed nearly all students in developmental psychology courses were female. It seemed reasonable these students were interested in child raising but that's not what they teach in developmental psychology.

Long before students enter college, they are taught that the subjects they study will not be of their own choosing, and that their job is to satisfy standards raised by various educational authorities for no apparent reason. Every subject they take in high school, for example, was selected for them by a committee in 1892 chaired by the President of Harvard. This committee wasn't concerned with what interested students but with making sure students wouldn't be unschooled in the basics so that the faculty would have an easier time teaching them.

It is the essence of our educational system that requirements and test scores matter and learning and ideas are secondary. This needs to change. Scholars from Aristotle, to Galileo, to Dewey, to Einstein have pointed out that real learning is by doing. The computer is a learning-by-doing device.

So, what about online courses? The attempt to replicate existing courses online is simply misguided. Online courses that seek to do more than present lectures and some assignments, those seeking to utilize the learn-by-doing power of the computer, to build the academic equivalent of the air-flight simulator and change the very nature of what is meant by a course, must accomplish at least one of the following:

  • Choose an area that is already doing-centered;
  • Choose a field in which there is no outside certification issue, where any change can occur if the faculty member wants to make that change; and
  • Utilize the expertise of a faculty member who is truly radical, one who believes students are not well served by the courses previously offered.

All this leads to three conclusions.

First, computer science is ripe for building high-quality online courses and curricula. Most computer science courses are doing-centered (or ought to be). Online computer science degrees are a reasonable idea and an idea we can do as well or better than the traditional kind. When lecturing is used to teach programming it is simply a misguided attempt at education. Programming is a trial-and-error affair; readily available online tutors and well-thought-out contexts are what matter.

Second, other kinds of skills, like writing, are also inherently doing-oriented and may be taught better online at least in part because there may be better instruction available online than could be found in one's local school. Any subject where a work product is produced is ripe for an online treatment. Written reports, complex calculations, analysis, and reasoning are the province of online education because help can be made available anytime, anywhere. Moving toward teaching subjects where students create works that can be criticized and improved is critical to effectively using online education and critical to helping to change our educational system to one of active learning-by-doing.

Third, subjects that do not have advocates in the undergraduate curriculum can be developed without faculty preconceptions about what must be covered, inhibiting their transformation into learning-by-doing courses. I am thinking of introductory medicine, law, and business, for example, intended for nonspecialists in those disciplines. Also, innovative master's programs intended to teach job skills are good candidates for online treatment. If courses present exciting virtual worlds in which to experiment, students will enroll in droves and thus demonstrate with their eyeballs and dollars they prefer these learn-by-doing/online courses. Savvy administrators will be able to increase enrollment by offering more similar courses, and perhaps the accreditation requirements will change.

The time has come for the Virtual University. The VU will have to be different than dear old traditional State U. I seriously doubt any university will be able to design effective online courses on its own any time soon. Universities won't be able to find the money to do the job right; they won't be able to convince the faculty to abandon cherished notions about what must constitute a course, and they won't be able to find the people who know how to design high-quality online courses.

There are professors everywhere radical enough to take a chance and design and offer courses students actually want. Such faculty may not all be at the same university, but then they don't have to be in the new Virtual University world do they?

Change is coming.

Back to Top


Roger Schank ( is Distinguished Career Professor in the School of Computer Science at Carnegie Mellon University.

©2001 ACM  0002-0782/01/1200  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2001 ACM, Inc.


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account
Article Contents: