Why is higher education more popular than ever? Are computer-literate students better programmers? Eric Roberts answers these questions and more in the following interview.
UBIQUITY: For the last six months you've been a visiting professor at Swarthmore. Tell us about that.
ERIC ROBERTS: I have an endowed faculty position. I'm the Eugene M. Lang Visiting Professor for Social Change, which is a wonderful job title. The position was established to bring people to Swarthmore College who are interested in all sorts of issues surrounding social change and the effect of various disciplines on society. And usually, it goes to a sociologist or political scientist, or someone like that.
UBIQUITY: Why did they choose you?
ROBERTS: Every now and again, they invite a scientist to look at science and social change and how those issues interrelate. The people who invited me here, Charles Kelemen in particular, have been asking me if I would be interested in doing this for quite some time. In the early '90s, I was president of the organization called Computer Professionals for Social Responsibility. From that, I did a lot of work with computing and its social implications. I've also taught at Stanford a course on computing, ethics and social technology since 1992.
UBIQUITY: Do you see the social issues related to information technology right now as reasonably benign? Or is there a state of urgency on particular issues?
ROBERTS: I think there's urgency on a number of issues. I think there are a lot of questions that have come up and have clearly become central policy questions. Not just those of us in the field realize this, but I think there's a lot of interest across the board. The Microsoft case, for example, is an important information technology question that has certainly gotten a lot of press this year. In my course here at Swarthmore, we used that as one of the focus study points that we looked at over the course of the last semester. People were trying to decide what the government should do, were they to have a favorable finding of law, which came out during the course of the semester. So people were anticipating and trying to think about what was going on in the real news events around them. So I think that the question of concentration of monopoly power in the IT industry is a very interesting and pressing one.
UBIQUITY: What are some of the other issues that you are concerned about?
ROBERTS: This whole notion of the digital divide that was described in the Commerce Department report two years ago and updated last year, for one. And then there is the whole nature of e-commerce, what it has meant for the economy and the shakeout that's going on now and the dot-coms. One of the issues that I've been most involved is the question of how the economic forces that surround the IT industry in this time of enormous excitement and great demand for IT professionals will play out in terms of the long-term future health of the information technology economy.
UBIQUITY: How do these economic forces affect academia?
ROBERTS: One of the issues that seems particularly problematic to me is the drain on academic computer scientists. This year, the available evidence suggests that there are only about one-third as many applicants for faculty positions as there are open faculty positions. This is unheard of in academia, where you advertise a job in history or English, and you get 200 applicants for each place. We only have one-third of an applicant for each place, which means that most of the positions will go unfilled; or at least will be filled by people who are not qualified, or who may have moved into those positions from other areas. It's a scary prospect. We had very similar situation in the early 1980s. Peter Denning wrote a well-known piece called "Eating Our Seed Corn," in which he described the process of industry desperate for talent raiding academic institutions to hire not just faculty, but also any students who might have thought about becoming future faculty. The short-term demands of industry indicate that they need to find the people, now, to maintain their competitive edge in this very competitive landscape. Very few students, say, at Stanford where I teach ordinarily, are going on past the Masters Degree to get a Ph.D. and to look at faculty positions, because it's just so exciting and lucrative to be part of the start-up world. I think we're looking to a time when, with enrollments rising in the field, departments are going to have to cut back their offerings because they won't have the necessary faculty.
UBIQUITY: What can be done about the problem?
ROBERTS: It's very hard to come up with solutions, because the economic forces that favor short-term gains by particular actors in the economy work against the long-term collective interest of that economy. So short term usually wins in those situations.
UBIQUITY: What about enrollment at the undergraduate level in computer sciences? Is that still very high?
ROBERTS: It's grown enormously in the past few years. One of the problems is that we don't have good recent data. The Department of Education statistics lag by almost five years in the surveys of how many people are graduating in each of the different fields. So, if you look at the statistics, what you find is this enormous run-up in the early '80s; peaking at about 1986; followed by a decline, which seems to bottom out in '92; and then a small rise past that for the next couple of years. But what the data don't show, because they're not current enough, is that every university has experienced an enormous increase in the last five years in enrollments in computer science courses. For instance, last year at Stanford, we graduated about 100 students with undergraduate degrees in computer science. This year, it was almost 150. So that's a one-year, 50-percent jump.
UBIQUITY: Is that a typical volume increase at other schools?
ROBERTS: I know that a lot of schools are reporting the same kind of increase. The number of students in our courses has doubled in the last four years, and the increase in the faculty size over that same period is only about 10 percent. So everyone is teaching a much larger class. The number of students per faculty member has gone up dramatically in that time, which makes teaching less attractive to educators. And so a number of schools are seeing significant attrition of their existing faculty.
UBIQUITY: You're working on a new curriculum. First, let me ask about the old curriculum. Are students, in your experience, happy with the state of academic computer science education?
ROBERTS: It's hard to know, because the surveys tend to be local; and much of the evidence anecdotal. My sense is that students at Stanford are very happy with the computer science education. We've worked on it very hard. We've taken student ideas into account when designing the curriculum. One of the things that I was hoping to do during the Curriculum 2001 project was bring some of the experience I have had, as Director of Undergraduate Studies at Stanford for the last ten years, to bear on that curriculum.
UBIQUITY: What is the purpose of the Curriculum 2001 project?
ROBERTS: What we're doing is offering curricular guidelines. The report is the next in a series that goes back a long time. The first major report was done in 1968 and was published by the ACM. Curriculum '68 was done shortly after the genesis of computer science as a field -- I mean, the first departments of computer science were formed only in the mid-'60s. An update was published in '78 and then another in 1991 that was done jointly with the IEEE Computer Society. We did some surveys in the early part of the Curriculum 2001 effort to look at the success of Curriculum '91, and found that in many places it had very little impact. It's actually easier to see evidence of the structure from Curriculum '78 in a number of institutions than it is to find Curriculum '91.
UBIQUITY: What's the reason for that?
ROBERTS: The basic problem was that Curriculum '91 was significantly more flexible and less, you could say, prescriptive. It provided less guidance for institutions that were looking for guidance. If you wanted to take essentially a turnkey curriculum and put it into operation, Curriculum '78 was pretty good. It gave sets of courses, and described those courses in some detail, and told faculty what they needed to cover; which also, by the way, made it easier for publishers to publish textbooks around those topics.
UBIQUITY: What was the approach in Curriculum '91?
ROBERTS: In Curriculum '91, there was a shift in focus toward providing a definition of the fundamental core of computer science that all undergraduates should know; done as a set of knowledge units, as they were called. This assemblage of knowledge units had to be manipulated to form courses. In order to provide flexibility, that was left largely to individual institutions. But very few individual institutions have the wherewithal or the time or the expertise, given the shortage of academic computer science talent, to turn those mix-and-match courses into real curricula. So in many cases, people just stuck with what they had. In Curriculum 2001 we're retaining this notion of a definition of a body of knowledge and using a similar structure to Curriculum '91 to categorize the discipline. But we are trying to provide much more guidance in terms of specific courses, and structures, strategies and tactics that individual institutions can adopt.
UBIQUITY: How far would you want to see that level of guidance go? Would you want to see a national curriculum, for example?
ROBERTS: No, I think that the idea is to make these curricular recommendations available, and make them easy to implement for institutions that want to take them as they stand. If an institution wants to modify it, more power to it. But to just say you must create your own curriculum leaves many institutions without the tools that they need to offer anything at all. What we want is something that people can adopt, that will have enough robustness that an institution will be able to adopt it fairly simply. We want to encourage institutions to extend it, to change it, to keep it alive, because computing is changing so rapidly that any static curriculum would almost certainly be out of date upon publication.
UBIQUITY: Right, that level of change is what I was just about to ask you about. Since things are changing now at Internet speed, is there any reason to examine the metaphor that underpins most of education? The metaphor of the course and the curriculum as a sequential process, a path of learning: Is it still a path?
ROBERTS: Well, it's never been exactly a path. There are usually multiple paths through a curriculum; it's not entirely a linear progression. There's a prerequisite structure that is still there. You have to learn to walk before you can run, even in the Internet analogies. I think we'll still see some of that structure. The question that a lot of people bring up is not that of the notion of sequentiality -- you start small and then move on to bigger and bigger tasks. The question is the mode of delivery: Does it make sense to have a lecture course in this time, when we have the Web and all of its enormous reach to provide material to a wide audience?
UBIQUITY: In view of all that, how do you respond to people who say that the Web will replace books and classrooms?
ROBERTS: I think that that's a strange view of what is education. I mean, we've had books since Guttenberg, and if someone wanted to know the material in a course they could pick up the book. It might be a little more static than a Web page, but it's not that different functionally. The institution provides a context for mentorship, so that someone can check how you're doing and encourage you. The notion that education is just about the transfer of knowledge leads to most of these misconceptions. Relatively little of what I do -- teaching the introductory computer science courses at Stanford -- is merely the transfer of knowledge
UBIQUITY: What do you do in the classroom that can't be taught on the Web?
ROBERTS: What I do is model the kind of behaviors that we want, so that people learn good engineering discipline, along with the knowledge to provide incentives and encouragement, and to get people excited about the field. It's much more psychological, and that's why that kind of education is so popular. I have people, even colleagues on my own faculty, who tell me that the university's days are limited. That this is all going to end soon. And I just turn around and say that's why education is more popular than ever before. I mean, we've had the Web for five years and during that time there's been just enormous increase in the popularity of higher education. The trend lines are wrong.
UBIQUITY: So why do some people have a doom-and-gloom scenario about education?
ROBERTS: I think that comes from a very narrow view of what is the function of education. People can take a book, and they can learn it, but they don't. People can take the Internet and learn it, but they don't. The way to get somebody to learn something is to provide them with the context, to provide them with the community in which that learning is validated, where there's the necessary motivation, where there's an excitement surrounding the learning. The Web has a certain level of excitement which, I expect, will wear off before too long. It doesn't have that human contact, which is so necessary to providing the kind of education that I value.
UBIQUITY: I'm curious about whether you've noticed a change in your computer science students over the years? Are they the same kinds of students you used to get, or are they different. Are they all dropping out to become millionaires?
ROBERTS: Well, at Stanford, a lot of them are. And I think that that phenomenon is true other places. Most of them don't drop out of college, although I have some advisees who have done just that. The majority of students leave at the end of their Bachelor's Degree or after they get a coordinated five-year Master's Degree. They don't go on for advanced degrees.
UBIQUITY: Why are students attracted to industry, other than the obvious financial rewards?
ROBERTS: The excitement of going into the start-up world is relatively new, within the last five to seven years. It is very much a pervasive attitude and view among the students that I know today. Even students who come in to a college with a more non-capital-based intellectual excitement about the field tend to be surrounded -- at least in our environment -- in this milieu in which everybody is pushing; not just the sense of making money through a start-up, but the sense of excitement that comes from that kind of a breakneck pace -- just working flat out to get something to exist. It's more of the cultural side of it. People are certainly flocking to that type of company, or that type of opportunity, much more than they did before.
UBIQUITY: Almost all students are computer-literate these days. Do you find they are better prepared for your course in computer programming?
ROBERTS: Students have used computers now for all their lives, and that's probably been true for a decade. I don't see much difference in the students' preparation for the introductory programming course. At Stanford, our introductory programming courses are taken by about 75 percent of the student body. So we have a wide distribution of majors and interest levels. In that cross-section of our undergraduate population, students do no better and no worse in the introductory course today than they did a decade ago, even though they've had more experience using the machine. There are many people who have real cognitive troubles understanding how to do programming. That has been true forever. And again, it's never the sort of thing that those people are going to solve completely on their own. They need to come up against problems that are too hard to solve for them, and fail, and get some guidance and feedback along the way for how they can do better the next time. That seems to be fairly much the same as it's been.
UBIQUITY: What do you think of the increasingly ubiquitous nature of computing; imbedded computing, and so forth. For example, MIT recently announced the Oxygen Project. Essentially, it's a research project devoted to making computers imbedded in everything.
ROBERTS: As prevalent as oxygen.
UBIQUITY: That's exactly right. My question was is what are the consequences of that for education? A frequent analogy that's made is between automobiles and computers . . .
ROBERTS: That millions of people know how to drive a car, but don't know what's going on under the hood.
ROBERTS: That's true today for computing, and it'll always be true for computing. But you need some automotive engineers. And you need much more in the way of software engineering.
UBIQUITY: Do you address this concern in that general course for undergraduates?
ROBERTS: Not exactly. The idea is that people gain something by being exposed to what is under the hood, to continue with the metaphor. There is a lot more intellectual rigor that goes on under that virtual hood that separates the hardware and the software world. I mean, when you look at the problem of writing programs that solve interesting problems, you get all sorts of wonderful intellectual challenges that I think are very liberating for a student mind, in a way that knowing exactly what's going on inside the carburetors and cylinders of the engine may be to some. But I think the course has a much broader application, partly because of the enormous flexibility of software and the fact that you can take those ideas and apply them to a wide range of problems. You're not looking at a single engineering artifact that solves one job, as you would be in the case of the automobile. You're looking at a set of intellectual tools and disciplines that will allow you to solve a wide variety of problems by applying those intellectually challenging ideas in the appropriate ways. And that's what a computer science course offers.
UBIQUITY: What are your thoughts on the future?
ROBERTS: It's clear that we will need more people trained than we have today with the growth of IT, even when it becomes imbedded. I think that's only going to increase the demand for people who are conversant with and talented in the creation of software for all of those different applications. This shortage of trained software people that we see today seems likely only to get worse, because we're going to need more and more people to create the software that runs the computer and the toaster and the shoe and whatever else you have.
A Ubiquity symposium is an organized debate around a proposition or point of view. It is a means to explore a complex issue from multiple perspectives. An early example of a symposium on teaching computer science appeared in Communications of the ACM (December 1989).
To organize a symposium, please read our guidelines.
Ubiquity Symposium: Big Data
- Big Data, Digitization, and Social Change (Opening Statement) by Jeffrey Johnson, Peter Denning, David Sousa-Rodrigues, Kemal A. Delic
- Big Data and the Attention Economy by Bernardo A. Huberman
- Big Data for Social Science Research by Mark Birkin
- Technology and Business Challenges of Big Data in the Digital Economy by Dave Penkler
- High Performance Synthetic Information Environments: An integrating architecture in the age of pervasive data and computing By Christopher L. Barrett, Jeffery Johnson, and Madhav Marathe
- Developing an Open Source "Big Data" Cognitive Computing Platform by Michael Kowolenko and Mladen Vouk
- When Good Machine Learning Leads to Bad Cyber Security by Tegjyot Singh Sethi and Mehmed Kantardzic
- Corporate Security is a Big Data Problem by Louisa Saunier and Kemal Delic
- Big Data: Business, technology, education, and science by Jeffrey Johnson, Luca Tesei, Marco Piangerelli, Emanuela Merelli, Riccardo Paci, Nenad Stojanovic, Paulo Leitão, José Barbosa, and Marco Amador
- Big Data or Big Brother? That is the question now (Closing Statement) by Jeffrey Johnson, Peter Denning, David Sousa-Rodrigues, Kemal A. Delic