Harvard announced Friday that up to 60 students have been forced to withdraw "for a time" as a result of a cheating scandal. Farhad Manjoo argued in September 2012 that the students should be rewarded, not disciplined. The article is reprinted below.
Last week, Harvard announced that it was investigating more than 100 students for cheating on the final exam of a course called “Introduction to Congress.” The class had a reputation for being easy, but many students found last spring’s open-book, take-home exam to be close to impossible. As a consequence, the kids apparently began to break the rules. Some students now face the possibility of being forced to take a one-year leave of absence from Harvard. A university official has called the case “unprecedented in anyone’s living memory.”
Certain students’ conduct does seem to have been indefensible—according to the Crimson, the exam-related malfeasance included plagiarism. But many of the accused did not copy their material. Instead, they merely worked with fellow students and their instructors to make sense of the tricky exam questions. What they did—work together to find an answer—should be encouraged. But too often in higher education, such collaboration is either given short shrift or actively penalized. Students are instead forced to find the answers on their own, in marked contrast to how they’ll be expected to behave once they graduate.
Students who spoke to the New York Times said that collaboration was widely thought to be allowed in the course. The class’s teaching fellows—graduate students who graded the exams and ran weekly discussion sessions—varied widely in how they prepared students for the exams, so it was common for students in different sections to share lecture notes and reading materials. The course’s instructor—Matthew B. Platt, an assistant professor of government—and the teaching fellows sometimes encouraged collaboration. During the final exam, some fellows even worked with students to define unfamiliar terms and help them figure out what, exactly, certain test questions were asking.
The test’s rules, though, explicitly prohibited such sharing. “The exam is completely open book, open note, open Internet, etc. However, in all other regards, this should fall under similar guidelines that apply to in-class exams,” the rules stated. “More specifically, students may not discuss the exam with others—this includes resident tutors, writing centers, etc.”
What’s the point of prohibiting students from working together? If the students in “Introduction to Congress” act as these test rules demand when they move into the workforce, they’ll be fired. Outside of academia, teamwork is the rule. Collaboration is widely hailed as a primary factor in creativity and problem solving. It’s the reason Pixar’s offices are designed to foster, in Steve Jobs’ words, “forced collisions of people” from different departments.
In this case, it’s the test’s design, rather than the students’ conduct, that we should criticize. In allowing students to consult a wide variety of sources, the Harvard exam was looking to assess something deeper than how well they could memorize and recall facts. Judging from some leaked questions, the test seemed to be designed to measure how students could think about some of the contradictions inherent in American government. (An essay question began, “Do interest groups make Congress more or less representative as an institution?”) But if you want to determine how well students think, why force them to think alone?
Outside of Harvard, these students won’t face many situations in which they’ll be prohibited from consulting with other people. Instead, they’ll have to act exactly as the alleged “cheaters” did in this case. Here, students who had poor teaching fellows sought help from peers who had better instructors. When they found exam questions that asked about material that hadn’t been covered in the course, they worked together to figure out how to answer them. (Some of them may have had no other choice, as Platt canceled his own office hours during the final exam.) Rather than punishing these students, shouldn’t we be praising them for solving these problems the only way they could?
Sure, the students’ collaborative work does make it difficult to assess individual performance—because many people’s answers sounded similar, instructors couldn’t determine who really understood the work and who was merely free-riding. Universities certainly still have an interest in measuring each student’s grasp of class material, and there are plenty of exam formats—in-class tests, timed essays, in-class discussions—that can do so. Open-book exams like the one for “Introduction to Congress,” though, are useful tools only when they’re collaborative. After all, the test allowed students to consult the Web, a medium that is built on teamwork. If students looked up stuff on Wikipedia or Quora, they would have been effectively discussing the exam with others. And yet online collaboration would have been kosher under the test’s rules, even though you’d be labeled a cheater if you posed the same question to your friend Laura rather than Quora. That distinction makes no sense.
I suspect this arbitrary distinction reflects universities’ discomfort with collaboration. Talking about an exam with your friends feels like cheating. But it’s time we realized it’s not, and that teaching people how to work together is a critical skill. Today, in most areas of life—the government, the military, science, the corporate world—real breakthroughs occur in groups. It’s time our universities prized group work, too.