Integrating an AI-Driven Discussion Platform: The Impact of Platform on Engagement and Quality

Abstract

Online educators rely on asynchronous discussions to satisfy the bulk of student-student engagement that is lost when transferring from a face-to-face to an online format. However, not all discussion platforms are created equal and some specialized tools may offer advantages over standard tools embedded in more comprehensive learning management systems (LMS). In this study, we compare two online discussion platforms – one is the native discussion tool embedded within the Canvas LMS and the other is a specialized discussion platform, Packback, which supports students and instructors with gamified elements and artificial intelligence. Specifically, we assess whether post quality differs across these platforms, as measured by average word count, cited sources and weekly participation rate. The initial results are mixed and differ by course, leading us to conclude that the impact of platform on engagement and quality likely depends on the course content, instructor and protocol.