Op-eds, Opinions

Çetintemel, Krishnamurthi: CS Growth and Student Support: Constraints and Challenges

By and
Op-Ed Contributors
Monday, February 3, 2020

In recent years, The Herald has published several articles and op-eds about growing enrollment in computer science, including suggestions on how to address the resulting challenges that students face. We appreciate this attention and care toward the student experience. As a department that takes pride in the deep engagement of undergraduates in its research and teaching programs, we welcome and value input from our students and appreciate constructive engagement.

We also understand the difficulties faced by our students. We sympathize: Our faculty members face similar problems from the other side! We, too, would love to teach small courses. We stay up at all hours dealing with student emails and issues. And as advisors, we feel the difficulty of offering personalized advice; there are currently more than 30 concentration advisees per faculty member on average.

Our department is not alone. All of our peer schools, from Harvard to Stanford, face similar demands for computer science classes and degrees. Each school has been struggling to meet this unprecedented surge and is experimenting with a variety of “fixes” that are consistent with their institutional resources, culture and values. On behalf of the department, we would like to enumerate some of our constraints as we work to accommodate the rapid growth in computer science concentrators.       

Hiring more faculty is certainly critical to addressing this increased demand. After an external departmental review in spring 2018, we worked closely with University leadership to develop a faculty expansion plan to increase the size of our faculty by 50 percent within five years — an average of about three new faculty members each year. This is an exceptionally ambitious growth plan for a Brown department! (We should note that the University determines the number of people we can hire in any year; the University rightly balances its needs and growth plans across numerous disciplines.)

But even with institutional commitment, growth is easier said than done in this climate. Every university that we compete with has numerous open positions in computer science, so it’s a seller’s market. In addition, top companies are also hiring the same kinds of people, extending the same terrific offers and perks that attract students to the discipline as undergraduates. Hiring quality people for the long term is extremely hard in this market, and we are fortunate to have hired outstanding faculty recently as it is. As a case in point, we spent much of spring 2019 conducting on-campus interviews, meeting 19 candidates and making several offers, to ultimately hire two new faculty members. Our peers report similar efforts and success rates.

Another “fix” considered by universities nationwide would restrict the size of the concentration. This is a fundamental design decision that our department currently faces: How do we balance the tradeoffs between access and an intimate feel? It would be easy for us to make classes small by just limiting access to the concentration or even to introductory courses. Several institutions have gone this route, either implicitly or explicitly; where explicit, it has often resulted in significant protests.  Either way, limiting the concentration size would likely significantly reduce opportunities for students who did not have prior exposure to computing, which would in turn greatly hurt the diversity of the student body. This would be especially unfortunate given the lack of diversity in computing — which has real, and unfortunate, consequences in the world. “We’re turning away many students we’d love to have. That’s the tragedy,” said University of Washington Professor of CS and Brown alum Ed Lazowska ’72 in an interview, as their department admits only a fraction of the students who intend to major in CS.

We have therefore made a conscious decision to be as inclusive as possible, even adding yet another introductory sequence this year to attract a wider variety of students and to reduce the size of the existing courses. The consequence of this increased access is more students in upper-level courses. Even when we execute our growth plan to completion, however, it is simply unclear how we could add enough faculty to return to class sizes of 30: With about 3,000 students enrolled per semester, that would require 100 faculty members!

As many readers know, we run a very large undergraduate TA (UTA) program. It has operated since 1965 and is a unique educational experience; other universities have studied and copied it, and many alums single out serving as a UTA as their most valuable experience at Brown. While we are proud of the program and all it offers, a larger and growing UTA program also means greater costs, greater quality variance and greater management effort. For decades, our UTA program has had a hierarchy that includes Head TAs and Meta TAs. These positions, which were primarily created to give students a sense of ownership and practice coordinating teams of people, have now become even more valuable in managing our large program. All things considered, we feel it would be a big disservice to significantly reduce the number of UTAs per student.   

We have also heard suggestions to rely more on graduate TAs. We agree that this would help improve the quality of CS undergraduate education. However, the number of graduate TAships is allocated by the graduate school, and we currently have a very limited number, though we have been requesting more. More graduate TAs would help — up to a point. In the end, graduate students are here primarily to do research; while there is of course much value to teaching, at some point too much time spent teaching can eclipse their career goals. Second, increasing the graduate program would also require increasing the number of faculty —  in essence, the size of our graduate program is proportional to the size of the faculty.

Under the circumstances, expecting any significant, timely improvements is unfortunately not realistic. We will continue to hire as many exceptionally talented faculty as we can. But even after this scheduled growth, we will still have many large courses. But we prefer this alternative to capping enrollments or limiting the number of concentrators in our department. Finally, our department is limited by the reality that our professors need to balance research and service (such as sitting on committees at Brown and around the world) with undergraduate education.

Ultimately, we hear and empathize with the concerns raised by students. Many of these are active topics of formal discussion in our faculty meetings and informal conversation in our hallways. We would love to engage with creative proposals that can ameliorate current conditions while taking into account the constraints we face. Please talk to us!

Professor and Chair of Computer Science Ugur Çetintemel and Professor of Computer Science Shriram Krishnamurthi can be reached at ugur_cetintemel@brown.edu and shriram_krishnamurthi@brown.edu. Please send responses to this opinion to
letters@browndailyherald.com and op-eds to opinions@browndailyherald.com.

Leave a Reply

Your email address will not be published. Required fields are marked *

*