A casual joke on Twitter recently let slip a dirty little secret of large science and engineering courses: Students routinely cheat on their homework, and professors often look the other way.
"Grading homework is so fast when they all cheat and use the illegal solutions manual," quipped Douglas Breault Jr., a teaching assistant in mechanical engineering at Tufts University. After all, if every answer is correct, the grader is left with little to do beyond writing an A at the top of the page and circling it. Mr. Breault, a first-year graduate student, ended his tweet by saying, "The profs tell me to ignore it."
While most students and professors seem to view cheating on examinations as a serious moral lapse, both groups appear more cavalier about dishonesty on homework. And technology has given students more tools than ever to find answers in unauthorized ways—whether downloading online solution manuals or instant-messaging friends for answers. The latest surveys by the Center for Academic Integrity found that 22 percent of students say they have cheated on a test or exam, but about twice as many—43 percent—have engaged in "unauthorized collaboration" on homework.
And cheating on an engineering problem set could be the perfect crime, in that it can be done without leaving a trace. Students in a large lecture course based on a best-selling textbook can often find the answer online, complete with all the math it took to get there.
How can a professor prove that the cheating students didn't work things out on their own?
Enter David E. Pritchard, a physics professor who teaches introductory courses at the Massachusetts Institute of Technology (when he's not in his laboratory devising new ways to use lasers to reveal the curious behavior of supercooled atoms).
Mr. Pritchard did detective work on his students worthy of a CSI episode. Because he uses an online homework system in his courses, he realized he could add a detection system to look for unusual behavior patterns. If a student took less than a minute to answer each of several complex questions and got them all right, for instance, the system flagged that as likely cheating. "Since one minute is insufficient time to read the problem and enter the several answers typically required, we infer that the quick-solver group is copying the answer from somewhere," he wrote in a paper last month in the free online journal Physical Review Special Topics—Physics Education Research.
He and his research team found about 50 percent more cheating than students reported in anonymous surveys over a period of four semesters. In the first year he did his hunting, about 11 percent of homework problems appeared to be copied.
Mr. Pritchard has no interest in becoming a homework cop. What he really wants to do is understand the minds of the offenders. The issue, he says, is far more nuanced than a story of "Top Students Caught Cheating." He told me that the dishonesty reveals flaws in the very way science is taught, and indicates an unhelpful spirit of "us versus them" between professors and students.
He believes that the most important part of learning physics comes by doing, and so students who outsource their homework learn little. His studies of his students prove his point. The cheaters generally perform far worse than other students come test time—students who frequently copied their homework scored two letter grades lower on comparable material on the exam.
Why Students Cheat
Here's what surprised me most when talking with people who have tracked college cheaters. Many students simply do not view copying homework answers as wrong—at least not when it is done with technology.
That's what Trevor Harding found. He's a professor of materials engineering at California Polytechnic State University at San Luis Obispo who has researched student cheating in engineering.
In surveys, he asked students if they viewed bringing a cheat sheet to an exam as cheating. Most did. Then he asked the same students whether they would consider it cheating to bring a graphing calculator with equations secretly stored on it. Many said no, that wasn't cheating.
"I call it 'technological detachment phenomenon,'" he told me recently. "As long as there's some technology between me and the action, then I'm not culpable for the action." By that logic, if someone else posted homework solutions online, what's wrong with downloading them?
The popularity of Web sites full of homework answers seems to confirm his finding. One of them, called Course Hero, boasts a free collection of "over 500,000 textbook solutions." The company set up a group on Facebook, where more than 265,000 people have signed up as "fans."
Drew Mondry, a junior at New England College, who recently transferred from Michigan State, is among them. "The feeling about homework is that it's really just busywork," he told me. (He said he does not cheat on his homework and only signed up as a fan of the Course Hero site because some friends did.) "You just call your friend and say, 'Hey, do you know the answer?'"
In the big science courses he has taken, professors didn't put much effort into teaching, so students don't put real effort into learning, he says: "I have yet to meet a professor who really loves teaching an introductory course, and that translates," was how he put it. "If you look bored out of your mind, guess how much I care?"
Some professors seem to believe that since students who cut corners on homework end up bombing exams, students get a kind of built-in punishment for the behavior, says Mr. Pritchard. Poorly performing students might even learn a lesson from their laziness. So the cheating will take care of itself, right? That's the rationalization, anyway.
Certainly, many professors put a lot of effort into their classes. And to them, blatant student cheating can feel like a personal insult. Eric Roberts, a computer-science professor at Stanford University who has studied academic cheating, told me about a student in his course who went to a public computer lab, found some other student's homework assignment saved on a machine there, changed the name to his own name, and turned it in as his own work. Except he left the other student's name on one page by mistake. Busted.
"This is lazy cheating," Mr. Roberts said. "They're trying to put one over on us. And if they're trying to match wits with us, I'd just as soon win—if that's their game I'll play it."
It isn't just professors who overlook cheating. Many colleges offer no comprehensive approach to minor academic cheating (the exceptions are institutions with honor codes, though few have them).
That's the view of Tracy Mitrano, Cornell University's director of information-technology policy. She recently attended a panel discussion on the campus where junior professors and students told stories of widespread cheating there—including a course where half the students routinely cheated on homework at least once a term. She started asking around and heard similar stories in several departments and at other institutions.
Now she's calling for universities to focus on the problem, and to devise a more standardized approach to punishing those who cheat on routine assignments. "Let's stop the cover-up of plagiarism," she wrote in an op-ed last month in Cornell's student newspaper. "The current system places too great a burden on individual faculty who would, under the circumstances, appear to have perverse incentives: Pursuing these matters lowers course evaluations, takes their severely limited time away from research for promotion, and unfortunately personalizes the issue when it is not personal at all, but a violation against the university."
Her proposal: Make it easier for professors to handle such cases, and reform academic judicial systems to make clearer distinctions between smaller violations, like homework copying, and larger ones, like cheating on exams. And assign appropriate punishments for each.
I ran that idea by W. Scott Lewis, president of the Association for Student Conduct Administration. "The short answer is she's right," he told me. "The system needs to be consistent. As it is, you might get caught cheating in physics, and that professor might say 'You failed my class.' Then you get caught cheating in an English class, and that professor says, 'You have to retake the final and get one letter grade lower.' That's inequitable." Even worse, the English professor and the physics professor will probably never talk to each other, so the serial cheater will never be reported to the institution's disciplinary system. In that case, the cheater wins.
In the humanities, professors have found technological tools to check for blatant copying on essays, and have caught so many culprits that the practice of running papers through plagiarism-detection services has become routine at many colleges. But that software is not suited to science-class assignments.
Mr. Pritchard, the MIT professor, did find a way to greatly reduce cheating on homework in his classes. He switched to a "studio" model of teaching, in which students sit in small groups working through tutorials on computers while professors and teaching assistants roam the room answering questions, rather than a traditional lecture. With lectures, he detected cheating on about 11 percent of homework problems, but now he detects copying on only about 3 percent.
It probably helps that he shares some crucial findings from his study with his students. Homework cheaters, he showed, are much more likely to get C's and D's on exams than those who work out the assignments on their own.