Frequency and cost of copying college homework revealed

According to research published online March 18 in Physical Review Special Topics: Physics Education Research, it turns out that unnoticed student cheating is a significant cause of course failure nationally.

A researcher from the University of Kansas has teamed up with colleagues from the Massachusetts Institute of Technology to get a better handle on copying in college in the 21st century.

Young-Jin Lee, assistant professor of educational technology at KU, and the Research in Learning, Assessing and Tutoring Effectively group at MIT spent four years seeing how many copied answers MIT students submitted to MasteringPhysics, an online homework tutoring system.

"MIT freshmen are required to take physics," said Lee. "Homework was given through a Web-based tutor that our group had developed. We analyzed when they logged in, when they logged out, what kind of problems they solved and what kinds of hints they used."

Lee said that it was easy to spot students who had obtained answers from classmates before completing the homework.

"We ran into very interesting students who could solve the problems -- very hard problems -- in less than one minute, without making any mistakes," said Lee.

Students also were asked to complete an anonymous survey about the frequency of their homework copying. (According to the survey, students nationally admit to engaging in more academic dishonesty than MIT students.)

Among the researchers' most notable findings:

• Students who procrastinated also copied more often. Those who started their homework three days ahead of deadline copied less than 10 percent of their problems, while those who dragged their feet until the last minute were repetitive copiers.

The students who copied frequently had about three times the chance of failing the course.

• Results of the survey show that students are twice as likely to copy on written homework than on online homework.
• This study showed that doing all the homework assigned is "a surer route to exam success" than a preexisting aptitude for physics.

"People believe that students copy because of their poor academic skills," Lee said. "But we found that repetitive copiers -- students who copy over 30 percent of their homework problems -- had enough knowledge, at least at the beginning of the semester. But they didn't put enough effort in. They didn't start their homework long enough ahead of time, as compared to noncopiers."

Because repetitive copiers don't adequately learn physics topics on which they copy the homework, Lee said, the research strongly implies that copying caused declining performance on analytic test problems later in the semester.

"Even though everyone knows not doing homework is bad for learning, no one knows how bad it is," said Lee. "Now we have a quantitative measurement. It could make an A student get B or even C."

At the beginning of a semester, the researchers found that copying was not as widespread as it was late in the semester.

"Obviously, the amount of copying was not so prevalent because the academic load was not as much at the beginning of the semester," said Lee. "In order to copy solutions, the students need to build their networks. They need to get to know each other so that they can ask for the answers."

But the KU researcher and his MIT colleagues also demonstrated that changes to college course formats -- such as breaking up large lecture classes into smaller "studio" classes, increasing interactions between teaching staff and students, changing the grading system -- could reduce student copying fourfold.