Should students use apps to write assignments? Attitudes on U of C campus surveyed

Can professors tell the difference between AI-generated writing and words written by a student? And could technologies such asChatGPT be used ethically for course work? Or would using them be akin to plagiarism?

These kinds of questions will soon be put to University of Calgary professors and their students as part of a research project led bySarah Elaine Eaton, an associate professor at the Werklund School of Education.

According to Eaton, who specializes in researching academic misconduct, there hasbeen a "knee-jerk" reaction by some educators, who have suggested banning students from using AI programs in their work.

But to Eaton, it's already too late to ban students from using the technology, especially college students.

"I think these apps are already everywhere," she said this weekon the Calgary Eyeopener. "We can't stop the train, so to speak. So the big question for us in higher ed isn't 'How do we ban this technology?' Instead, it's 'How do we use it ethically and how do we teach our students to use it ethically?'"

Eaton's research group is composed of researchers from several departments, such as English and engineering, and includes an official withthe university'sstudent accessibility services.

"We all have really different starting points," Eaton said. "In education, I'm concerned about how this is used in the classroom. My colleague in English is wondering howthis is used in creative writing, and my colleague in engineering is really interested in machine learning he says bring it on."

The results of their study could help shape academic misconduct policies at universities across Canada, if not the world, especially as the public increasingly comes to realizehow powerful the AI technology is, and how rapidly it's improving.

Did AI eat your homework?

Late last year, the San Francisco-based company OpenAI released its ChatGPTprogram to the public. The cutting-edge chatbot left users from journalists to technologists, educators to businesspeople astonished and perplexed.

Given virtually any prompt, the large language model could quickly produce a textual response with surprising eloquence and accuracy.

At first glance, the technology posed a significant problem for any class that assesses writing, particularly for students in elementary and secondary education.

However, Jason Wiens, a professor in the English department at the University of Calgary, who is part of the AI research group, saidChatGPT isn't yet producing university-level writing.

Starting this month, the research group will begin collecting data, surveying students and educators on the University of Calgary campus.

The questions will probe how well students and professors understand AI technologies and measure the extent to which each group can differentiate between AI-generated text and writing created by a human.

Additionally, and perhaps most importantly, the surveys will ask students and professors their thoughts on ethical questions surrounding AI use in the classroom and on assignments.

The research grouphopes to have results by the end of the term.

Learning from AI

Educators are already thinking of creative ways to use AI for teaching purposes.

For instance, Wiens suggested that in an assignment focused on editing text, students could be tasked to give an AI a prompt and then edit the resulting draft, revising and improving the language.

He noted that while AI is getting better at generating text, similar technology could also be used in the future to critique or even grade it.

"In an absurd scenario, you could have a student submitting work generated by a machine and then have it evaluated or assessed by a machine," Wiens said, adding this isn't the academic future he necessarily hopes to see.

To Eaton, the technology is something educators will need to grapple with, because these programs will be part of the landscape in which students will build their careers.

"I think it's kind of incumbent on us to learn how to teach students ethically, because chances are that they are going to have access to these apps when they go to work," she said.

On the question of misconduct, Eaton said she's unaware of any AI provisions on the books in any Canadian academicmisconduct policy.

"This tech wasn't even around a year ago," she said. "So, for people to automatically say it's misconduct or automatically say it's plagiarism, I urge caution on that, because I'm not convinced it is."