80 suspected of cheating
By Annabelle Boyd
Between 80 and 90 students in Introduction to Computers and Engineering Problem Solving (1.00) have submitted "duplicate code" on problem sets, according to course professor Nigel H. Wilson SM '70. This is the largest instance of student cheating uncovered in MIT's recent history.
Most of the code plagiarism involved small groups of students who handed in exactly the same computer program code or very similar coded statements on their problem sets, Wilson said.
At the beginning of the term, Wilson went to each of the course's 18 recitations and informed the students what was expected of them individually and what constituted cheating.
"I told the class that while I would like every student to solve every problem by himself, I recognized this was unlikely. If a student got stuck, it was appropriate for him to speak to a TA, another student or myself. However, jointly written code and <>
the submission of jointly written code were unacceptable."
Wilson first became aware of the plagiarism in late April when a student came to his office and expressed "great frustration" about the cheating which, this student felt, was unjustly raising the class average.
After the student left, Wilson contacted several TA's and discovered that many of them felt they "may have seen similar code on the problem sets." Taken with the "large" discrepancy between test scores and problem sets, this sentiment indicated to Wilson that there was a problem.
On April 27, Wilson called a meeting with his teaching assistants and a plan was devised to test for similarities in the problem sets that had been received from the previous week. Wilson asked one of his TA's to write a computer program which would screen for similar statements in the student code.
The problem sets were then run through the program, and the ones that triggered were "carefully and individually" analyzed for evidence of copying by the teaching assistants and Wilson.
"We were shocked at the number of times duplicate code appeared," Wilson commented. "If someone had come to me the month before and told me what I would find, I wouldn't have believed them."
Other series of problem sets were then tested, with similar results. "Some of the students had copied code on up to five different problem sets, while others had done it on only one," Wilson said.
"It has taken a long time for us to go through all of the problem sets and to locate all of the duplicated code," he added.
On Monday, April 30, Wilson announced to his class that cheating had been discovered, and that any student who had duplicated code would receive a zero for the corresponding problem set.
Later, Wilson also decided to forward the name of each student found cheating to the Committee on Discipline for further review.
In a few days, when Wilson has the final list of those students who submitted duplicate code, he will mail them each a "confidential" letter informing them of the number of zeros they have received on problem sets, and of the fact that their name has been submitted to the COD.
According to Wilson, some students in his lecture have been visibly upset by the cheating, and by the fact that they do not yet know if their name is on the list. Others have complained that the timing of the allegations was unfortunate, since it happened so shortly after drop date.
"Morale in the class is understandably low," he said.
Wilson, who has taught 1.00 three times, does not feel that the cheating discovered this term was particularly unprecedented. "I would be surprised if this problem is unique to this specific class and this specific term," he said.