Students are feeling a lot of concern and skepticism about generative AI, Ecampus researchers found in a recent survey, and the results indicated a need for some level of guidance around the use of generative AI such as ChatGPT or Copilot.
The Ecampus Research Unit is using these findings to develop practical recommendations for Ecampus course designers and instructors and has shared the full survey and results. The unit will also be repeating the survey this fall, as so much has changed around AI uses in the past year.
The anonymous survey, completed by 669 students who took an Ecampus course in fall 2023, asked undergraduate, post-baccalaureate and graduate students about their perceptions of generative AI tools. However, most results were derived from the responses of the 546 participants who said they had heard of generative AI.
Generative AI is a subset of artificial intelligence referring to systems capable of generating content, including text, images and computer code.
As of fall 2023, the majority of survey respondents (60%) said they had not used generative AI in their Ecampus courses. A similar majority (55%) said they had not used generative AI for professional activities, but 53% said they had used it for personal activities (hobbies, household or family use).
The overwhelming majority said ChatGPT was the generative AI tool they used the most, and their most common classroom and professional uses were for brainstorming or generating ideas, explaining difficult concepts and learning new content or skills.
On a question about whether students were interested in integrating AI into Ecampus coursework, the results found an almost perfect three-way split between “yes,” “no” and “unsure.”
“To a certain degree, we were surprised by just how wide the range of responses was,” said Mary Ellen Dello Stritto, director of the Ecampus Research Unit. “What we were most excited about was the emotions question, which we hadn’t seen in other surveys.”
When the survey asked participants to select how generative AI made them feel from a list of eight emotions — “curious,” “anxious,” “confused” etc. — more than 15% of students (79) selected “other” and wrote in their own, resulting in 54 unique emotion words in addition to the survey’s list. From the original eight, the most common emotions were “curious” and “concerned;” among students who wrote in their own, the most frequently listed were “angry” and “disappointed.”
Students had concrete examples when describing negative perceptions of AI, including mis- and disinformation, inaccuracy, copyright infringement, job loss and information bias, among others.
They were also concerned by the inconsistency of AI policies in their online classes, though there were significant differences between undergraduate, post-baccalaureate and graduate student responses.
Participants’ responses were more vague when they spoke of positive possibilities, such as help with work or help getting a job.
Ecampus has developed and is continuing to develop a resource guide for the use of generative AI Tools.
The study provided researchers with a trove of qualitative data in response to the more open-ended questions, so the Ecampus Research Unit will be working on several white papers in the coming months to flesh out what students are thinking and feeling about AI.
Related Stories
