Next-Gen Predictive Models for Student Support and Policy

Wednesday, October 16, 2019 | 4:00PM–4:45PM CT | Board 109, Poster Area, Halls F1, F2, Level 3
Session Type: Poster Session
Delivery Format: Poster Session
This presentation will focus on a novel approach consisting of identifying composite student behavioral profiles that capture actual performance and attendance patterns, which are then used in predictive models to forecast behavioral profiles of new students. A discussion of the benefits and limitations of this approach will follow.

Outcomes: Understand the limitation of predictive models forecasting binary outcomes * Review how cluster analysis can be used to first develop a rich understanding of the actual performance and life cycle of students * Learn how membership in these behavioral profiles can be forecast for entering students

Presenters

  • Loraine Devos

    Director of Institutional Effectiveness, The University of Arizona Global Campus

  • Stephen Nettles

    AVP of Assessment and Evaluation, The University of Arizona Global Campus