In a world where collaborative processes drive the most successful enterprises, most Stanford professors still test the knowledge, skills and abilities of their students by the same archaic methods used for centuries. This editorial board believes that the time has come for a reinvention of the way in which we assess student performance, starting with exams.
In January of 2010, Provost John Etchemendy and then-Vice Provost John Bravman appointed the Study of Undergraduate Education at Stanford (SUES) to reevaluate the University’s approach to undergraduate education. The last major study, in 1994, yielded fundamental reforms: freshman seminars, increased opportunities for undergraduate research and revisions to undergraduate writing and IHUM programs. In their charge to the committee this time around, Etchemendy and Bravman wrote, “[the] growing social, political, economic, and ecological interconnectedness of the world certainly challenges us to look more broadly at what it means to be an educated citizen.” They ask the committee to investigate how such global changes, over the past 15 years, have affected what students need from their undergraduate education and, accordingly, how Stanford should adjust the way it educates. Within SUES, the Subcommittee on Student Learning is asked to address the question of how to encourage exploration and sustained reflection among Stanford students. In order to best answer this question, SUES should scrutinize the closed-book exam as a means of testing student preparedness for an increasingly connected and collaborative world.
In today’s global economy, the lone worker is increasingly a relic of the past. Programmers rely on the Internet to find example code, companies crowd-source ideas and peers are always available for consultation via phone or e-mail. In an interview with Harvard Business Review, Cisco CEO John Chambers argued that a new emphasis on collaboration and teamwork has allowed his company to innovate at an unprecedented pace. SUES has the goal of preparing undergraduates for “local, national and global citizenship.” Given this charge, is the same method of testing a student in isolation an accurate measure of what is needed to compete in the workforce?
Furthermore, it is easy to see that low-value problem sets and high-value exams motivate students to relax throughout the quarter and place immense pressure on themselves to cram during the last days of school. If Stanford’s goal is to promote gradual exploration of material, rather than the simple ability to commit knowledge to short-term memory, then it should task SUES to thoroughly investigate whether the current emphasis of most classes on exams is appropriate.
Consider a typical student enrolled in the Physics 20 series. If she spends 5-10 hours per week on problem sets, then by the end of the quarter she will have spent 75 hours per quarter on work that is worth 18 to 20 percent of her final grade. To spend a proportional amount of time on a final exam worth 40 percent, that same student would have to spend 10 hours a day for more than two weeks — 150 hours — studying for a single exam. Since this is impossible, exams are worth a disproportionately large amount.
One obvious counter-argument is that time spent on problem sets can be double-counted as time studying for finals, and students’ performance on the final exam should reflect the effort invested in problem sets. Yet, even with this generous assumption, professors would be hard-pressed to explain why they trust the evaluative rigor of a single data point — one three-hour exam — more than twice as much as 10 weeks’ worth of gradual and sustained learning.
An argument in favor of the current system is that it provides an even playing field for evaluating students. More simply, everyone is equal when faced with a blue book, but some students may use the Internet or friends to gain unfair advantage on homework. However, exams do not level the playing field so much as they subject all students to the same absurd metric. Test-taking ability can be unrepresentative of understanding and a poor predictor of future success. Those who argue that closed exams are a true test of an individual in isolation have already made the flawed assumption that individuals ever work in isolation in the real world.
On the SUES web site, Etchemendy and Bravman noted that the 1994 review of undergraduate education led to reforms that “increased the rigor, coherence and clarity of [the] undergraduate program, while engaging faculty and undergraduates with one another more deeply than perhaps ever before.” If the current study hopes to be equally effective, it must consider the changing demands of the 21st century on our graduates and adapt teaching methods accordingly. In particular, SUES should question the validity of exams, which so disproportionately shape the Stanford experience. The Faculty Senate and the Committee on Undergraduate Standards and Policies would do well to heed the results of such a study.
Doing away with exams should not be taken as a decline in the hard-nosed work ethic of yesteryear or an assault on meritocracy. Instead, any revision to the way in which Stanford grades its students should encourage more effective learning and prepare students for the kinds of work encountered in the workplace. In doing so, it may save the next generation of Stanford students much anxiety, while elevating Stanford as not only a center of innovative discovery, but an institution of innovative learning.