Amid increasing expectations for socioeconomic relevance, higher education confronts, in many countries, a similar set of challenges: declining general-support levels linked with more performance-based funding, expanded enrollment demand, an increasingly knowledge-based and global economy, and a more intense managerialism. While giving unprecedented centrality to academic work, deteriorating conditions of work and of increased accountability has placed more performance pressure on the faculty.
Student Retention and Attrition guidelines are part of the Federal Government’s performance based funding framework. One of the recommendations from the Higher Education Standards Panel review is to consider changing students’ enrolment prior to census date when a certain level of engagement is not met. This study investigates this recommendation by trialing and testing a model to see if completely disengaged students are able to be retrospectively identified as at risk of failing all subjects. Using learning analytics alone to create a predictive model at scale proved to be very difficult. When applied to session 1 of 2019, even the strictest criteria included five false positives out of 17 identified students. There is promise, however, that a hybrid model of learning analytics with additional oversight from teaching staff could be a solution, but this needs further research.
Discipline and Institution Commitment: Professorial Views
