By MARTY LEVINE
“Can artificial intelligence improve the learning experience for the next generation of physical therapy students?” asks Reivian Berrios Barillas, assistant professor in the School of Health and Rehabilitation Sciences’ Department of Physical Therapy, in a proposal that won one of this year’s Momentum Awards at Pitt.
The answer, Berrios Barillas believes, is yes — particularly in the human anatomy course that she teaches to beginning PT students here, and in similar courses taught in all health sciences professions.
“Can faculty in PT’s human anatomy course identify who is at risk for poor performance and why?” the proposal further asks. What is the amount of time students spend on the various curricular resources in this class — the readings, videos, interactive exercise, laboratory sessions and quizzes? Which items do high-performing students access versus low-performing students? What specific feedback can the instructors suggest to specific low-performing students to improve their individual outcomes?
The proposal aims to test an online dashboard for this partially online course — to track student progress and point them toward areas they should likely review, based on their individual performance so far — and to test a new chatbot to supplement students’ ability to consult the instructor.
This is a foundational course in the program, Berrios Barillas points out, and success here helps dictate success in later courses. Thus, for such an important course, AI can be an on-the-spot, timely tutor for students.
The new chatbot will be using only material culled from previous iterations of the course — not information skimmed from the Internet. The chat of this chatbot will be based on needs students have already demonstrated in past semesters, and on the instructor-created materials that helped previous classes.
“A nice thing about the chatbot, it does answer sophisticatedly,” says Berrios Barillas, since it has learned what course instructors know.
This fall, her team will be piloting the chatbot and dashboard. “Catching where students are not doing well in the course and being able to help them sooner” is the overall goal, she says. “It helps them remediate before they don’t do well in the course.”
To create these new course aids, she is teaming with faculty in the School of Computing and Information, the Dietrich School of Arts & Sciences and the School of Education.
SCI faculty member Peter Brusilovsky, director of its Intelligent Systems Program, brings his experience with digital textbooks and assessing student engagement. Scott Fraundorf, a psychology faculty member and research scientist in the Learning Research and Development Center, is a cognitive psychologist concerned with the metacognition of learning, while Education faculty member Lindsay Clare Matsumura, associate director of the LRDC, also brings her insight on why students learn the way they do, as well as on the most effective feedback methods for assessing students. And Kim Nixon-Cave, program director of the PT department’s hybrid PT doctorate option, lends her lengthy experience in PT curriculum development.
Online learning in general has received a mixed review from students, Berrios Barillas notes. Student perceptions of AI as a tool for learning also can be a hurdle. But with some predicting that demand for physical therapy professionals is expected to grow 15 percent in the coming years — a faster pace than many other job categories — there will be increased need for hybrid human anatomy courses that can make sure every student is mastering the material.
This proposal hopes to demonstrate that AI is a workable and valuable add-on to such courses — here and elsewhere. Thus, the SCI team is focused on creating the AI mechanisms to assess and improve students’ learning and performance, while the psychology/education team will examine how these mechanisms are faring among students, with an eye toward improving them.
Pitt Momentum awards are designed to move projects to a next level, where national funding can be sought. “I can see it used throughout our entire program,” Berrios Barillas says of these new assessment and assistance tools — as well as in other SHRS programs and in other schools. After this fall’s pilot phase, in the spring the team will refine its methods and publish a how-to, so others can begin to adapt these methods, and then pursue greater funding.
“We’re just really excited to get it moving,” she says. “We think it will change the way teaching is provided” to students — and make the job easier for faculty.
Marty Levine is a staff writer for the University Times. Reach him at martyl@pitt.edu or 412-758-4859.
Have a story idea or news to share? Share it with the University Times.