Students often have difficulty estimating how well they know a topic, which can lead to inefficient learning or suboptimal educational outcomes. A new project led by Associate Professor Dong Wang and Assistant Professor Nigel Bosch aims to improve students' ability to estimate their knowledge using artificial intelligence (AI) methods. The researchers were recently awarded a three-year, $850,000 grant from the National Science Foundation (NSF) for their project, "A Metacognitive Calibration Intervention Powered by Fair and Private Machine Learning."
"Students in college are often expected to do a considerable amount of studying and learning outside of class hours, especially in online courses, which requires a high degree of self-regulation and metacognitive knowledge to study effectively," said Bosch. "However, there are few opportunities for specifically learning self-regulation and metacognitive skills, especially early on in courses, while there is still time to improve studying skills in advance of major assessments (like finals)."
"While there is a rich set of research on AI methods in educational contexts, those efforts rarely consider some of the key social and human factors, such as privacy and fairness, that are needed for widespread adoption of personalized educational software," added Wang. "This project addresses these issues with a novel decentralized AI framework that is specifically for education contexts."
For their project, the researchers will utilize the predictive power of machine learning to anticipate how well undergraduate students will perform in a course. Then, they will teach the students to recognize their trajectory while there is still time to improve it.
"For example, we might find after a few weeks of a course that we can predict a student will probably get around a C+ on an upcoming test, whereas the student might think they are on track for an A," said Bosch. "We will provide students with some exercises to self-assess and improve their ability to estimate their own learning, so that they can better prioritize and motivate their studying strategies."
The AI systems being developed will not directly access student data, in order to reduce biases related to key aspects of students' identity. By improving AI "fairness" in this privacy-focused situation, information about students cannot be directly used to audit or adjust the models. According to the researchers, the privacy and fairness capabilities of the project framework will transform postsecondary online learning.
"This project will advance AI research by incorporating, for the first time, both a strict privacy guarantee for student data and fairness considerations across multiple student demographic groups," said Wang. "It will also advance education research by determining how effective preemptive feedback is for improving knowledge estimation skills and examining the mechanism by which this estimation influences academic outcomes."
Wang's research interests lie in the areas of human-centered AI, social sensing and intelligence, big data analytics, misinformation detection, and human cyber-physical systems. He holds a PhD in computer science from the University of Illinois Urbana-Champaign.
Bosch holds a joint appointment in the Department of Educational Psychology in the College of Education. His research focuses primarily on machine learning and human-computer interaction applications in education. He holds a PhD in computer science from the University of Notre Dame.