WPI Receives Grant for Development of Software Tools to Enhance Student Learning
Worcester Polytechnic Institute (WPI) has received a one-year, $277,044 grant from the Bill & Melinda Gates Foundation to develop software tools that can automatically detect whether students are engaged or disengaged while using educational software, and identify specific types of disengagement—information that can be used to improve learning outcomes. The project is one of several investments by the foundation aimed at dramatically increasing the number of “college ready” students who will graduate from high school by 2025.
"We are delighted that the Gates Foundation has selected WPI to participate in this important program," said WPI President and CEO Dennis Berkey. "The foundation's mission to significantly improve the quality of learning and the level of student success in high school and beyond aligns well with WPI’s multifaceted efforts to augment teaching and learning in K-12 science, technology, engineering, and mathematics. Like the Gates Foundation, we recognize that keeping students engaged in these subjects and motivating them to continue on to careers in science and technology is vital to our nation's competitiveness."
Ryan S.J.d. Baker, professor of psychology and the learning sciences at WPI, is principal investigator for the project. He and his team will develop “automated detectors of learner engagement,” which are programs that can be integrated with educational software. As part of the project, detectors will be developed for at least four specific online learning systems that will be selected by WPI in cooperation with the Gates Foundation. The detectors, in real time, will record and analyze data about how students interact with the educational software looking for telltale patterns that indicate the student's degree of engagement as well behaviors that correlate with disengagement, such as boredom, off-task behavior, systematic guessing, and over use of hints. Extensive field observations in schools of students using the software as part of their regular classroom instruction will be used to verify that the detectors accurately identify engagement and disengaged behaviors. Since the detectors work without the need for any additional technology, Baker says they can easily be scaled to thousands or even millions of students.
Baker said that being able to accurately measure student engagement and intervene when students are disengaged will not only result in short-term gains in learning, but can have critical long-term effects on students' academic performance and careers.
“Poorer performance on tests, poorer learning gains and outcomes, lack of motivation in learning, hostility and dissatisfaction toward school—these are all associated with disengagement," he said. "Off-task behavior is also associated with more serious behaviors later on, such as skipping class or dropping out of high school. On the other hand, engaged concentration not only leads to significantly better learning gains, but also is a major contributor to factors that predict long-term career success."
The project builds on Baker’s extensive research, sponsored mostly by the National Science Foundation, on the interactions that take place between students and educational software, work that has led to software that adapts to individual differences. To study engagement, robust learning, and emotion in real classrooms, his research combines quantitative field observations of student behavior while using educational software with data mining to detect patterns in the ways students tackle the tasks that the software presents. His studies of gaming-the-system and off-task behavior have influenced the design of educational software used worldwide.
An internationally-recognized expert in educational data mining and learning analytics, he was recently elected the inaugural president of the International Educational Data Mining Society. He was chair of the third International Conference on Educational Data Mining in 2010, was co-editor of the Handbook of Educational Data Mining (2010), and is currently associate editor of the Journal of Educational Data Mining. He is the director of WPI's Educational Psychology Laboratory and a member of the Pittsburgh Science of Learning Center.
Because education beyond high school vastly improves an individual’s opportunities for success, the Gates Foundation has set a strategic goal of having 80 percent of all high school students in the class of 2025 graduate ready for college, without the need of remediation. To achieve this goal, the foundation is seeking to understand how to maximize learning, recognizing that engagement is a critical piece of the puzzle.
Recognizing that current methods for measuring engagement—self reports and teacher reports on surveys, observations by researchers, and video analysis— are subjective, obtrusive, and inconsistent, the foundation has launched a new program to develop new measurement methods that can improve teaching and learning. In addition to the work led by Baker at WPI, the foundation is funding teams at the Harvard-MIT Division of Health Sciences and Technology, which will conduct laboratory studies of engagement, and Clemson University, which will conduct field studies of engagement. These teams are working with MIT and Affectiva, which will lead engineering and analytic efforts related to measuring engagement..