USING LMS AND ASSESSMENT DATA TO IDENTIFY UNDER-ENGAGED STUDENTS AND PREDICT PERFORMANCE

Abstract

There is a wealth of data and metadata that students produce simply by being enrolled in our Units of Study and being subject to our learning management system (LMS), learning activities, and assessments, and of this, even the simplest metadata may have relative predictive value for students’ continued engagement with the unit and eventual performance. Across 2023, our School made a concerted effort to collate and analyse this data and metadata from all three year-levels of our undergraduate Psychology program, with the aim of identifying ‘under-engaged’ students, then using our Student Relationship Engagement System (SRES) to personally contact these students ahead of the semester’s census date with advice, access to support services, and encouraging them to make informed choices about their continued candidature. As a result of this, we have a within-cohort grouping variable of ‘engaged’ versus ‘under-engaged’ with which we can run post hoc analyses on performance once the semester has ended. In this presentation, I will report data from our School’s efforts, suggest what level of predictive value (if any) there are for different kinds of LMS and assessment data, and provide actions that individual academics, ‘third space’ staff, and/or relevant professional teams could implement in their unit(s) or program to intervene with support services or even merely identify ‘under-engaged’ students for exploratory analyses in their reviews of unit(s) or program.