SBIR Phase II: Intelligent tutoring system with EEG-based instructional strategy optimization

This SBIR Phase II project will develop a computer-based tutoring platform that adapts its teaching strategy to each student by monitoring their brain activity. Access to individualized instruction by teachers or qualified tutors is limited for a vast number of students, particularly in disadvantaged areas and for science, technology, engineering and mathematics (STEM) subjects. Computer-based teaching technology could meet this demand, but currently relies on indirect inferences of a student's state, and therefore lacks some of the key abilities of human teachers to assess a student during a lesson. This project will address this challenge by enabling the tutoring system to determine how hard students are thinking and their level of focus. This will be accomplished by measuring a student's brain electrical activity using a newly-developed headset designed specifically for school-aged children. Information derived from brain activity patterns will be used to guide tutorial topic selection, difficulty of the content and the level of interactivity. This project could lead to a significant advance in education technology by enhancing the adaptability of computer-based tutoring platforms and expanding the depth of information on student performance available to teachers and parents. Adaptive computer-based education is now a rapidly growing market, and the technology developed in this project aims to provide companies with a unique capability to improve their products and to meet the need for truly scalable individualized tutoring. The unique innovation of this project is to create a tutoring platform that closes the loop with the student by adapting its teaching strategy to the student?s cognitive state. To overcome the limitations of current adaptive tutoring systems, this product will measure the student's cognitive workload and engagement, and use this information to adapt aspects of the instructional strategy such as topic selection, content difficulty and interactivity in a closed-loop fashion. The goals of this project are to modify current brain-wave measurement systems for use by children and young adults, extend associated computational algorithms and validate them to an industry standard, and evaluate the effect of adding brain-based metrics to student models that are at the core of adaptive technologies. To meet these objectives, we will perform research and development to advance technologies for brain-activity monitoring, and conduct studies with students in grades 6 through 12 who will be tutored in subjects such as math, biology and history while their brain activity is recorded. The results will be used to determine the effects of using brain activity data to improve the adaptability of computer-based tutors and expand the depth of information available to educational data analytics platforms.