By MARTY LEVINE
Peter Brusilovsky is aiming to create a new generation of textbook — both interactive and adaptive to each user’s learning abilities and pace — to teach one of the more popular but difficult subjects in the School for Computing and Information: Python programming.
His project — one of the seven winners of this year’s Innovation in Education Awards from the Provost’s Advisory Council on Instructional Excellence — will bring together two streams of his research, including work that won Brusilovsky, an SCI faculty member, a previous Innovation award.
Interactive electronic textbooks already exist, which have links among content for easier navigation and to videos for another mode of learning besides reading. But Brusilovsky says this technology has not been combined with adaptive textbooks, which can both assess how well users undertake Python programming exercises and point them toward their individualized next steps within the textbook — whether that next step is on thus-far unread pages or involves a specific review.
Python programming is very popular and the school has many courses in it, but “it is one of the most challenging subjects” to learn, he says. “And having interactive tools is important. There are a lot of students and they need a lot of support.”
He thus plans to develop AdaPT, “which fuses a well-tested electronic textbook with a large volume of smart interactive learning content: worked code examples, code animations and several types of automatically assessed practice problems,” his proposal explains. “The system uses advanced learner modeling to trace the growth of learner’s knowledge and skills and to identify misunderstood concepts.”
This summer, he and his project team interviewed fellow Python instructors to see what smart content would be best to use in a textbook geared to different types of Python courses and uses. The new interactive and adaptive textbook will be piloted this fall
“There’s a lot of smart interactive content for Python,” he notes. Artificial intelligence “systems can tell if you are right or wrong and guide you through many steps to get there. But it has never been used with a textbook. We have realized: together they can be the best.”
Before now, interactive textbooks have only been able to chart how much has been read, assuming each reader is ready to move onward once the last word of a section has passed before a student’s eyes. This is not only a poor way to judge whether knowledge has been absorbed, but it also is not at all geared to each learner individually.
On the other hand, an adaptive textbook can use AI to assess a user’s understanding of concepts (as each demonstrates through problem-solving) and then guide students individually to the next textbook lesson they are ready to work on — or back to what they apparently didn’t absorb yet.
“We realized that students need more practice — engaged in doing things and getting feedback from the system,” Brusilovsky says. “Reading is important for learning but it is not the way we really learn knowledge.” Interactive problem solving is most effective, he notes. “And then we can guide you to the next part of the book that is best to read for you.”
Quick learners will be guided through the textbook at a more rapid pace and slower learners will be guided “back and forth,” he says. This method ought to particularly help those whose background did not include the most rigorous high schools or programming courses.
About a decade ago, Brusilovsky received a National Science Foundation grant to develop a better electronic textbook — work that helped prepare for this project. He hopes for further NSF funding to move this project beyond the Innovation award’s prototype development stage.
Once the new textbook is in use, he says. “I hope that learning will be more engaging” for individual students. He also hopes this textbook design can be a resource for students and faculty who undertake the mastery of Python programming on their own.
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.