Michelle Callaway
Manager, Program Review,
San Jacinto College District
Leveraging Software for Predictive Analytics
Tuesday, March 06 | 3:30PM–5:00PM ET | Orlando Ballroom M-N, Convention Level
Session Type:
General Session
Higher education is turning to analytics to drive decision making around student success initiatives. This section of the summit will include presentations from three institutions that have used analytics in support of student success:
● "The Role of Business Intelligence and Data Science in Student Success," Western Governors University
● "Using Data to Promote Student Success and Departmental Accountability," University of South Dakota
● "Leveraging Software for Predictive Analytics," San Jacinto College
(STUDENT SUCCESS) Leveraging Software for Predictive Analytics: Accountability in higher education requires that institutions demonstrate student success. San Jacinto College is using predictive analytics to help increase student success by identifying students who are at greatest risk of dropping out from one semester to the next. The college has built a homegrown system to develop predictive models that can be leveraged to increase student persistence. We will discuss how we identified the student population on which to focus, built our predictive models, chose the most appropriate models, and presented our results to key stakeholders. Topics of discussion for future work will include adding noncognitive data to the models.
● "The Role of Business Intelligence and Data Science in Student Success," Western Governors University
● "Using Data to Promote Student Success and Departmental Accountability," University of South Dakota
● "Leveraging Software for Predictive Analytics," San Jacinto College
(STUDENT SUCCESS) Leveraging Software for Predictive Analytics: Accountability in higher education requires that institutions demonstrate student success. San Jacinto College is using predictive analytics to help increase student success by identifying students who are at greatest risk of dropping out from one semester to the next. The college has built a homegrown system to develop predictive models that can be leveraged to increase student persistence. We will discuss how we identified the student population on which to focus, built our predictive models, chose the most appropriate models, and presented our results to key stakeholders. Topics of discussion for future work will include adding noncognitive data to the models.
