Please refer to Part 1 for the six essential considerations for addressing AI-driven cheating. Part 2 discusses how you can redesign assignments using the TRUST
HIGHER ED TEACHING STRATEGIES FROM MAGNA PUBLICATIONS
The launch of the artificial intelligence (AI) large language model ChatGPT was met with both enthusiasm (“Wow! This tool can write as well as humans”)
Plagiarism seems like a clear-cut crime: if the words of another author appear in one’s writing without appropriate attribution, that writer has “stolen” those words. U.S. higher education institutions take the offense seriously: minor cases often result in probation, suspension, or expulsion. This black-and-white perspective toward plagiarism, however, does not effectively identify, prevent, or resolve writing issues.
The most common approach to cheating involves trying to prevent it—multiple versions of a test, roving observation during tests, software that detects plagiarism, policies that prohibit it. However, if we look at cheating across the board, what we’re doing to stop it hasn’t been all that successful. Depending on the study, the percentage of students who say they’ve cheated runs between 50% and 90% with more results falling on the high side of that range. Can we be doing more? Here are some ideas.
Although there are software-based services that can help instructors check the originality of student writing and discourage students from deliberately copying the work of others, many instances of plagiarism stem not from a willful disregard of the rules but from simple ignorance of them.
Elizabeth Kleinfeld, an English instructor and director of the writing center at Metropolitan State University of Denver, has studied plagiarism and students’ use of sources for the last seven years, mostly among students in first-year writing courses. She has found that many students don’t understand the differences between paraphrasing, summarizing, and plagiarism.
Although some behaviors are pretty much universally identified as cheating (copying exam answers, for example), we’re not in agreement on everything. Particularly significant are disagreements between faculty and students (for example, students don’t think cheating occurs if they look something up on their phone and can’t find it; faculty consider cheating in terms of intent). In many cases, there is the question of degree (when, for example, collaboration crosses the line and becomes cheating). The effectiveness of cheating prevention mechanisms can be increased by clarifying upfront what is and isn’t cheating. Here’s a collection of activities faculty can use to ensure that students understand the behaviors that constitute cheating.
I’m still wandering around in the literature on cheating. It’s hard not to get depressed. It’s such a pervasive problem and one that compromises all that education could and should be.
Faculty are pretty much focused on preventative measures, which are essential, but there are a couple of other issues rarely mentioned in the literature or in our discussions. Students who don’t cheat usually aren’t on our side when it comes to enforcing cheating policies. In one study, almost 93% of the students said they had witnessed another student cheat, but only 4.4% said they had ever reported a cheating incident (Bernardi, et. al., 2016) Students are in a bind—they don’t want to rat out fellow classmates, some of whom may be friends. If they do and word gets out, they are labeled as “snitches” and “tattletales” — told to mind their own business and otherwise berated. With serious social consequences like these, it takes real courage to do the right thing.
Cheating among college students remains rampant. Our institutional and/or course policies aren’t stopping much of it. There are lots of reasons why, which we could debate, but the more profitable conversation is how we get students to realize that cheating hurts them. I don’t think they consider the personal consequences, so that’s the goal of this memo, framed like others that have appeared in the blog. You are welcome to revise it, make the language your own, and share it as you see fit with students. Will it stop cheating? Not likely, but it might make some students realize the consequences go well beyond getting caught.
As online education continues to grow, so does the potential for academic dishonesty. So how do you ensure your online students are not cheating on their tests? Bottom line, you don’t. But there are ways to stack the deck in your favor.
The good news is it’s not as bad as you think. A 2002 study by Grijalva, Kerkvliet, and Nowell it found that “academic dishonesty in a single online class is no more prevalent than in traditional classrooms” (Paullet, Chawdhry, Douglas & Pinchot, 2016, pg. 46). Although the offenders have become quite creative in their endeavors, the prevention remains the best defense.
When I first I started teaching, I knew what plagiarism meant and how it related to schoolwork. But student “cheaters” challenged my beliefs. I also assumed graduate students would submit original work. So it took me by surprise when I noticed a mysterious improvement in one student’s writing capacity, well beyond the skill level he’d demonstrated earlier. When a Google search proved more than 20 percent of his paper was copied, he explained it as a computer error—he’d accidentally dropped the footnote when cutting and pasting. I lowered his course grade, but assumed it really was a snafu—not subterfuge. The (now) obvious question went unasked: Why was so much of his assignment based on other people’s insights?