More accessibility, more problems? The rise of online learning has made it easier for more people to participate in education. No more snow days. Students learn from home. No more long commutes. College students have the opportunity to work from their dining room table.
For many people, this is a positive development. Not only was it of enormous importance during the pandemic, but it continues to be an excellent tool for people who are balancing other responsibilities in addition to school.
Unfortunately, we all know that the internet comes with downsides. It’s easy to cheat or commit fraud when no one else is in the room with you.
How can cyber experts prevent cheating on online classes or exams?
First, What Are the Risk Factors?
Cheating in school is nothing new. It’s only that online classes make it easier than ever. Here are a few of the ways that sneaky students can exploit e-learning vulnerabilities:
- AI Plagiarism– A relatively new but certainly ongoing concern. Students both remote and traditional can relatively easily generate work that is not their own and try to pass it off. While most colleges have anti-AI policies, it has been very difficult to enforce. Yes, the AI checker came back positive. No, that does not necessarily mean that the student used AI. It could just mean they are a mediocre writer.
- Google– It’s easy to use search engines to answer test questions during remote exams. While there are ways to keep a student from using Google on the same computer they are taking the test on, what’s to stop them from using their phone or tablet to look up answers?
- Outright Fraud– Finally, how hard is it to simply pay someone else to do the work for you? There are people all over the world who are willing to take remote classes on behalf of someone else for a fee. Obviously, this is a price-restrictive way to cheat, but people do it every day.
Unfortunately, none of these problems are completely solveable. Universities develop new security protocols. Cheaters find novel ways to exploit the same old problems. But while there aren’t many perfect solutions out there, it’s certainly possible to reduce the chances of online cheating.
The AI Problem: Develop Different Criteria
Right around the time ChatGPT was first released, a college professor made headlines for failing his entire class on the grounds of AI plagiarism. This story, which played out at Texas A&M, was unfortunately the first of many such dilemmas.
Worried about the potential for cheating, the now infamous teacher wanted a way to screen his student’s work for evidence of AI. His solution? Go directly to the source.
Dr. Jared Mumm uploaded his student’s essays into ChatGPT, and asked what probably felt like a reasonable question: Did you write this?
ChatGPT, in its infinite wisdom, said yes, for every single essay.
The problem, of course, is that the application was wrong.
There are AI checkers that could have produced a more accurate response, but even those are imperfect. A positive AI score does not indicate that AI was actually used. It just means that patterns typical of artificial intelligence were detected in the text.
Often, bland or otherwise mediocre writing is flagged incorrectly. Technical writing—where a complicated topic is broken down in academic language, might also come up positive. You know, like the sort of language a student might use on a college essay?
Because plagiarism is the sort of accusation that can end an academic career, AI checkers simply aren’t an adequate way of enforcing AI bans.
What is a professor to do?
Colleges can reduce the odds of AI plagiarism by developing rubrics that AI would have a hard time following.
This could involve elaborate in-text citations. AI is typically pretty bad at sorting real sources. It could also involve including oral presentations or simple conversations as a way of ensuring that the student actually understands the topic they have been writing about.
Another thing to keep in mind? For now, at least, AI isn’t capable of writing A or B college papers. The writing is mediocre. The insights, often surface-level. The accuracy, pretty dismal.
So while it is hard to ensure kids don’t cheat with AI, it is still easy enough to make sure that they produce high-quality work.
Timers
Another imperfect but potentially impactful technique. Test timers make it harder for students to endlessly Google answers. They still might give it a try, but a ticking clock will at least encourage students to focus on studying rather than relying on search engines.
Depending on the program being used, you may even be able to time individual questions, making it even harder to cheat using the internet.
Visually Verify Student Identities
Having students keep their cameras on while they are taking a test might feel like an invasion of privacy. In reality, it’s no different than what would be expected of a student in a traditional learning environment.
This requirement is already in place for many standardized assessments, like the NCLEX or the LSAT. Mandating it can reduce, or even eliminate the potential of academic identity fraud.
An Imperfect Solution?
Well, sure. But remember—even with traditional learning, there is no perfect way to keep students from cheating. They can write answers on their hand. Look things up on their phone. Copy off the person sitting next to them.
Teachers can design ways to thwart these methods, but the determined cheater will simply develop a new approach. Academic dishonesty is nothing new. While virtual learning creates new challenges, a refined approach can go a long way toward preventing cheating.
Indeed, technology in virtual settings offers both obstacles and opportunities in combating academic dishonesty. Innovative solutions like software that monitors keystrokes and eye movement during exams can be part of this refined approach. Ultimately, fostering an educational culture that emphasizes integrity and the value of learning itself may be the most effective deterrent against cheating.
About the Author
Ryan Ayers is a researcher and consultant within multiple industries including information technology, blockchain and business development. Always up for a challenge, Ayers enjoys working with startups as well as Fortune 500 companies. When not at work, Ayers loves reading science fiction novels and watching the LA Clippers.
Leave a Reply