On Declining Interest in CS and What Can Be Done…


A few interesting pieces/tidbits to juxtapose this morning. Sam Liles helpfully forwarded this piece from The Tennessean on the declining interest in computer science as a major, which is apparently getting a fair bit of play on digg.com. The article asks the now familiar question:

Computer science majors make some of the highest starting salaries for college graduates in the country, at about $50,000 a year. Computer science and computer engineering jobs are some of the fastest-growing occupations in the nation, according to the U.S. Department of Labor.
So why are university computer science departments watching their enrollments slide?

The article puts the finger on student’s perceptions about the state of the job market — that potential majors shy away from CS because of fears about offshore outsourcing. But it also does an “ok” job of showing how that might be a mistaken impression:

The East South Central region, which includes Tennessee, Kentucky, Alabama and Mississippi, is the fastest growing in the country in terms of information technology jobs, in part because of economic growth here, according to her agency’s latest survey.
Some 23 percent of chief information officers in that region plan to hire more workers this year and only 1 percent plan cutbacks.
Movva said she hasn’t been able to find experienced consultants in Nashville, and has had to hire outside the region, including signing visas for foreign nationals, to fill job openings.
“There are lots of jobs but not enough people are entering this field,” said Sandeep Walia, who is opening an e-commerce software office called Ignify on West End Avenue.
With Oracle database experts making as much as $150,000 a year, “you wonder why more people aren’t getting into this,” Walia said.
Vanderbilt professors are worried about the perception that jobs aren’t out there.
The department’s Web site includes a plea from the chairman to prospective students that says: “Contrary to what you may be reading in some publications, there are jobs. …
“The jobs are out there, but the perception is that they’re not,” said Richard Detmer, the chairman of the computer science department at Middle Tennessee State University.
Jonathan Waite graduated with a bachelor’s degree from Vanderbilt in May. But he says the job market is saturated with computer scientists. He feels that way even though he got three job offers in three months of looking for a job.

But students’ perceptions of the job market aren’t the only aspects of the problem worth addressing. Increasingly, CS departments are realizing that the way they teach computer science might have something to do with declining interest in their major, too. And that’s the focus of this piece in today’s Inside Higher Ed, “New ‘Threads’ for Computer Science.” The piece (which must be good because it quotes my boss, Andy Bernat, and CRA Board Member Rich DeMillo) focuses on the announcement of planned curriculum changes in the College of Computing at Georgia Tech, where DeMillo is Dean.

The Georgia Institute of Technology is today unveiling what some experts believe is a much broader approach to the problem. The institute has abolished the core curriculum for computer science undergraduates — a series of courses in hardware and software design, electrical engineering and mathematics. These courses, in various forms, have been the backbone of the computer science curriculum not just at Georgia Tech but at most institutions.
In their place, Georgia Tech is introducing a curriculum called Threads.

Underlying this approach is the view that “the one size fits all approach to computer science just isn’t working anymore,” said Richard A. DeMillo, dean of the College of Computing at Georgia Tech. The plans were developed by professors, who prepared a white paper outlying how this approach would create “symphonic thinking” graduates — another way of saying graduates whose jobs wouldn’t be outsourced, a fear keeping many out of the field.
“The really big change here is that we were willing to give up the idea of a core curriculum,” said DeMillo. “If you have 90 percent of your courses occupied with the core, you don’t have the flexibility to do anything creative.”

The Georgia Tech approach is noteworthy, not just because it’s an interesting approach to the problem, but because — as Andy points out in the article — it’s being undertaken by one of the bigger schools in computing. There’s plenty of additional detail on Georgia Tech’s approach in the article and on the Georgia Tech website.
Additional efforts in improving the quality of CS education will likely be give a boost by NSF’s very recent solicitation for its new CISE “Pathways to Revitalized Undergraduate Computing Education” (CPATH) program. The new program will make $6 million in awards in FY 2007 to encourage “colleges and universities to work together, and with other stakeholders in undergraduate computing education including industry, professional societies and other types of organizations, to formulate and implement plans to revitalize undergraduate computing education in the United States.”
While the image of computing still requires a lot of work, it’s also becoming increasingly clear that the field needs to reexamine the way it educates its undergraduates. In the coming months, I think we’ll see further efforts by the various computing societies (including CRA) to put a focus on CS education. Hopefully the NSF solicitation will uncover some interesting ideas and approaches within the discipline as well.

On Declining Interest in CS and What Can Be Done…