Learning Science in Service to Students: Applying Artificial Intelligence and Automatic Question Generation to Scale Learn by Doing


Educational technology has undergone rapid changes in recent years as artificial intelligence has begun to shape how data and content are used in student learning resources. However, as new technology is inserted into the learning ecosystem, it is paramount to ensure it is based on learning science research and created with a student-centered learning engineering framework. Student success should remain at the heart of all educational technology, and even more so when artificial intelligence is applied. In this study, we will describe how artificial intelligence is harnessed for automatic question generation in order to provide students with formative practice questions while they read textbook content. This learn by doing method creates the doer effect—a learning science principle studied at Carnegie Mellon University and proven to cause better learning. Yet this learn by doing method is often limited by the availability of resources to create and implement practice questions in digital learning resources. To use artificial intelligence to drastically increase the availability of this learning science method harnesses technology in a student-centered approach. Furthermore, the learning engineering framework not only serves in the creation of educational technology, but encourages results to be iterated on and shared with the research community. We will evaluate the performance of these automatically generated questions using student data from courses. These findings will provide insight into the learning benefit of formative practice generated through artificial intelligence and directions for future research.

Author Information
Rachel Van Campenhout, VitalSource, United States
Benny Johnson, VitalSource, United States

Paper Information
Conference: ERI2022
Stream: Artificial Intelligence and Adaptive Learning

The full paper is not available for this title

Video Presentation

Comments & Feedback

Place a comment using your LinkedIn profile


Share on activity feed

Powered by WP LinkPress

Share this Research

Posted by James Alexander Gordon