CS Department remains atop Honor Code breaches November 28, 2011 1 Comment Share tweet Mary Harrison By: Mary Harrison Computer Science (CS) continues to contribute the largest percentage of Honor Code violations among all University departments, although the proportion of plagiarism cases coming from CS is declining. From 1991 to 2001, the CS Department accounted for, on average, 37 percent of plagiarism cases. Between 2006 and 2009, that proportion dropped to 25 percent. However, the real average increased from 14 to 27 cases per year, which coincided with a spike in CS enrollment. (SERENITY NGUYEN/The Stanford Daily) “The fraction of [plagiarism] cases that have come from CS has gone down from those years [the 1990s] when in some years, we were well above 50 percent of total incidents,” said Eric Roberts, professor of computer science and a member of the University’s Internal Review Panel, which will be submitting new recommendations about academic integrity at Stanford early winter quarter. In 2002, Roberts published a paper about plagiarism in the Stanford CS Department, which asserted that between 3 and 5 percent of CS assignments were plagiarized at the time it was written. “It doesn’t feel nearly as out of control as it did when I was writing that paper,” he said. There are several theories as to why CS accounts for such a large proportion of honor code violations, among them real-world incentives and more rigorous regulation. “I think a good piece of it comes from the economic incentive,” Roberts said. “People think if they get a CS degree they could be the next Larry Page or Sergey Brin. To cheat on a philosophy paper doesn’t have the same economic incentive attached to it.” Roberts also named the availability of copy-able code, the fact that many CS courses recycle assignments from previous years and the cumulative nature of CS classes as incentive and opportunity to cheat. CS professor Alex Aiken developed the department’s plagiarism detection system, known as Measure of Software Similarity (MOSS), which is used at universities around the world. He rejected the possibility that the higher levels of plagiarism in CS classes come from a heavier workload. “It is fairly well known that CS classes are difficult and involve a fairly heavy workload, but I don’t think it’s that significantly different from, say, a humanities course where you have to write big papers,” he said. Judicial Affairs, however, has evidence showing that workload is a major contributor to plagiarism. “Most often students reported running out of time, a heavy workload or concerns about their final grade as reasons for attempting to take a shortcut,” wrote Judicial Affairs Officer Jamie Pontius-Hogan in an email to The Daily. CS is one of the few departments that uses a plagiarism detection system on a regular basis. Aiken said that when schools adopt the MOSS system, “they find a whole bunch of cases right at the start, word gets out and then it drops to a low level quite quickly.” He also clarified that MOSS “doesn’t find plagiarism, that’s a fairly common misconception–it finds similarity.” “Plagiarism has an intent,” he said. “Similarity is a more technical notion. When we find things that look the same, we don’t know necessarily why they look the same. Humans still have to look at it.” Many departments choose not to check for plagiarism because of Stanford’s Honor Code, which leaves the responsibility for honesty in the hands of students. “Faculty [in other departments] often catch plagiarism because the text does not read like the students’ previous work or appears to be written at a different level,” Pontius-Hogan said. Parna Sengupta, associate director of the Introduction to the Humanities (IHUM) Program, which averaged seven instances of plagiarism per year between 2006 and 2009, said that most faculty rarely check for plagiarism. “In general, we expect students to uphold the Honor Code,” she said. The CS professors who spoke to The Daily, however, are split on how much influence the Honor Code actually has on behavior. “Data show that schools that have an honor code have less incidents of cheating than those without honor codes,” Roberts said. Aiken, on the other hand, is not sure if this data is genuine. “The thing I’ve noticed about honor-code schools is because the penalty is so high, typically expulsion, there is a reluctance on the part of faculty to prosecute plagiarism because they sometimes think punishment is not proportional to crime,” he said. He did add that he thought it was getting easier for professors to report instances of plagiarism to the judicial committee, because “there are definitely degrees of infraction [affected by] intent and the degree to which things were copied.” Roberts agreed, saying that intent is what defines plagiarism. “What happens if someone is trying to cite something on an assignment and does it wrong?” he asked. “In my mind, that is not an Honor Code violation.” “What I think is most important is that the student recognize that the Honor Code is theirs,” Roberts continued. “Student lobbied for it…to make sure they could be in charge of administering academic integrity.” Alex Aiken Eric Roberts Honor Code judicial affairs MOSS Stanford Computer Science 2011-11-28 Mary Harrison November 28, 2011 1 Comment Share tweet Subscribe Click here to subscribe to our daily newsletter of top headlines.