The grant says the research will look at how children communicate STEM content, how they learn to persist through challenging work, and how teachers support and promote noncognitive skills, often called soft skills, like persistence.
The overall goal of the project is to find ways to use artificial intelligence to help students who might be struggling with three skills that support learning: persistence, academic resilience and collaboration. Part of the project is building a database of how kids learn, using 96,000 youths across 24 school districts, according to the University of Illinois at Urbana-Champaign.
Eventually, the AI tools can work with teachers in classrooms to support students in developmentally appropriate ways, according to the grant.
“One of the biggest obstacles to AI advances in learning is the lack of shareable datasets for K-12 students,” said Cheng-Xiang Zhai, a computer science professor at Illinois. “We will remedy this by collecting large representative datasets with rich contextual information about learner interactions to enable foundational advances in AI areas, including fair and robust machine learning, natural language processing and socially intelligent agents.”
Schroeder said the research literature has so far been focused on AI and learning outcomes, but the focus on this project is to support those skills to help students succeed.
“One of my primary research areas on this project will be working with our collaborators to explore how to design agents that support these skills and self-beliefs,” Schroeder said.
Schroeder said he would be working in part to create characters that would pop up and interact with the student.
“For example, a student might be learning about math and get something wrong while working their way through a problem,” said Schroeder. “At that point, a virtual character might pop up on the screen and help provide them an explanation of what they did wrong in the problem. Maybe it will ask them a question that will cause them to self-reflect on why they chose one answer rather than another.”
Schroeder’s background is in designing computer-based learning environments and has designed virtual characters before.
Schroeder said he is excited to work on such a large and comprehensive project and to have an opportunity to collaborate on such a large scale.
“I find it really exciting that we are bringing together all of these different skill sets to design intelligent systems to support noncognitive skills,” he said. “Opportunities like this don’t happen every day.”
About the Author