The Council—which adjudicates academic dishonesty at the College—heard 128 cases in total, a slight increase from the 115 it heard last year. The uptick in inappropriate collaboration comes the same year that a cheating scandal in Harvard’s flagship introductory computer science course swept campus: More than 10 percent of 2016 fall enrollees in Computer Science 50: “Introduction to Computer Science I” appeared before the Council last semester.
In its report, the Honor Council obliquely referenced CS50, writing in a footnote that “one large introductory course” had skewed the data for last year.
“In 2016-2017, the Honor Council received a significant number of reports from one large introductory course,” the footnote reads. “The Council works to protect the privacy of students and does not report on cases by individual courses, but the impact of the reports from this one course is evident in the statistics that follow.”
The footnote mentioned that the number of cases per class year, as well as the number of cases per academic discipline, were the two “areas of measurement” most “significantly impacted” by the unnamed introductory course.
The number of freshmen appearing before the Council doubled from the previous year, rising from roughly 24 to 49. In the 2015-2016 academic year, then-sophomores—members of the Class of 2018—were most likely to be investigated. This past year, though, then-freshmen—members of the Class of 2020—were most likely to be investigated.
CS50 was 53 percent freshmen in fall 2016, according to data provided by course head instructor David J. Malan ’99.
For the second year in a row, the vast majority of academic dishonesty cases occurred in courses offered in the Sciences Division or by the School of Engineering and Applied Sciences. CS50 falls under the the latter category.
In an interview Friday, Honor Council secretary Brett Flehinger declined to comment on whether the marked increases in inappropriate collaboration cases and in freshman cases were due to cheating in CS50 last fall, citing concerns over “confidentiality” and student privacy.
But he suggested looking at the footnote about the introductory course in particular when trying to interpret the Council’s academic dishonesty data.
“We don’t talk about individual courses, but we do have a footnote in there that says we got a lot of reports from one large introductory course, and you can draw the conclusions that you will,” he said. “One of the reasons we put that in there is that it does shift the stats, [so] we’re open about that.”
Following the cheating scandal, some CS50 students charged that the course’s collaboration policy for problem sets—summarized as “be reasonable”—was too vague and led to unintentional violations by enrollees. The Council’s final report particularly focused on academic dishonesty cases arising from “problem sets and collaboration issues” and offered three recommendations to rectify the issue.
The Council proposed that Harvard consider an “educational campaign” to remind students about collaboration policies, that faculty “revise and improve” their collaboration policies, and that the College “refine” the language it offers to faculty on collaboration. Malan made a number of changes to CS50 this year, including requiring enrollees to attend two orientation meetings where course staff offered advice on how to navigate issues of academic integrity.
Dean of Undergraduate Education Jay M. Harris attended at least one of these sessions and implored students not to cheat, asserting that “no one should leave this room not understanding 100 percent what it is that you may and may not do in this course.”
Out of all cases reviewed by the Council in 2016-2017, the body found that roughly three-quarters of students investigated had violated the College’s Honor Code—roughly the same percentage as in 2015-2016. In total, 47 percent of all students who appeared before the Council were found responsible for an academic violation that resulted in a “change in status,” with punitive measures ranging from academic probation to temporary withdrawal from the College. In 2015-2016, 44 percent of those investigated suffered a change in status.
Twenty-four students—roughly nineteen percent of all cases—were asked to take time away from Harvard in 2016-2017, an increase from the 14 students and 12 percent of cases who were asked to do so last year.
Flehinger declined to answer a question asking how many of the students asked to withdraw were CS50 enrollees.
“The minute a course gets named, any student who is involved in that course… immediately thinks the comment is aimed at them,” he said.
—Staff writer Hannah Natanson can be reached at firstname.lastname@example.org. Follow her on Twitter @hannah_natanson.
More than 60 Fall CS50 Enrollees Faced Academic Dishonesty Charges
Under CS50 Policy, Accused Students Likely Left in Dark
With CS50 Cheating Cases, FAS Dean Taking Hands-Off Approach
In ‘Unusual’ Timeline, Honor Council Review of CS50 Cases Lasted Through February
Some Accused of Cheating in CS50 Sought Legal Counsel