The Promise and Peril of Predictive Analytics in Higher Education: A - PowerPoint PPT Presentation

The Promise and Peril of Predictive Analytics in Higher Education: A Landscape Analysis Manuela Ekowo Policy Analyst New America January 6, 2017

Enrollment Management What it is Who is using it Admissions teams use past student demographic Wichita State University and performance data to -- Wichita, Kansas make predictions about if prospective students are likely to become an applicant, be admitted, and enroll at the institution.

Adaptive Technologies What it is Who is using it Digital courseware programmed to use and Glendale Community store data on how students College interact with the tool in --Glendale, Arizona order to direct when the tool should display certain course content to a student to help them gain mastery, and how and when a student should be assessed on whether they understood course content.

Early-Alert and Course Recommender Systems What they are Who is using them Early-alert : Student demographic and Integrated Planning and performance data are used Advising for Student to flag which students may Success (IPASS) grantees be at-risk of failing a course or dropping out of Austin Peay State school altogether. University Course recommender : Can -- Clarksville,Tennessee use this same data to suggest majors students should pursue and courses they should take next.

Challenges to Using Predictive Analytics Ethically • Labeling and Stigma • Transparency • Privacy and Security

Minimize Labeling and Stigma Don’t close off students’ futures or profile students traditionally at-risk. Source : Kent Weakley, Shutterstock.

Ensure Transparency Have transparent tools, processes, and uses. Source : Lars Hallstrom, Shutterstock.

Guarantee Privacy and Security Have guidelines and policies that address who has access to student data and predictive results, how students will be informed about the institution’s data use practices, and how to ensure data is secured in all locations. Source : Maxx-Studio, Shutterstock.

Road Map to Use Predictive Analytics Ethically • Define a common vision and plan • Build a supportive infrastructure • Ensure proper use of data • Design predictive models and algorithms that avoid bias • Carefully deploy interventions

Questions, Comments, Concerns?

Thank you! Manuela Ekowo Policy Analyst New America ekowo@newamerica.org @ekowohighered http://www.newamerica.org/education-policy