The power of nudge: Altering choice defaults based on student risk to improve student outcomes

Background/context: Tertiary education access is expected to grow. However, some students come to university and not only fail, but arguably fail with little to no benefit, receiving ‘zero-fails’ in which students do not submit any assessments and receive a cumulative mark of zero. The sector should seek to understand and reduce the incidence of zero-fails. Altering choice architecture (or nudging) has been an effective method to influence behaviour in other contexts.
Description: In 2022 and 2023, students at census who identified as at risk of zero-fail grades informed by learning analytics had their choice architecture changed: they were required to opt-back-in if they wanted to continue.
Method(s): Using an action research design, learning analytics (including grades, LMS access and assessment submission) were used to identify students at-risk of zero-fail grades. 596 students had their choice architecture changed and were contacted. Differences in outcomes between course levels and commencing status were compared using z-scores.
Evidence: 35 percent of students did not reply and were unenrolled due to the changed choice architecture. 37 percent of students replied and chose to unenroll. 27 percent opted-in to continue studying. Of the students who chose to continue, most failed more than half of their units with 82 students receiving at least one zero-fail. There was no statistical difference between response rates or decisions made by students split by commencing status or course level. Postgraduate students were more likely to pass all units when they chose to continue compared to undergraduate students.
Contribution: Selective alteration of the choice architecture based on learning analytics can improve outcomes for high-risk students by reducing unnecessary debt and zero-fail.