Notes from “Creating a Framework for an Institutional Data-Driven Approach to Student Success” presented by Elizabeth Mulherrin, Assistant Assistant Vice Provost for Student Success and Jack Neill, Senior Director of Data Analysis, both from University of Maryland University College, at the 2015 ELI Annual Meeting.
Evolution in how we use data
- Institutional – external reporting, internal metrics (usually owned by institutional research offices)
- Transactional – tracking behavior in LMS, flag risk groups in CRM and SIS
- Predictive – build models based on data with historical and behavioral data
Legacy data systems require you to “pull data out” and look back in time, instead of offering real-time insights.
Dashboards provides real-time information, see trends, compare to other departments, institutions, etc. Push vs. Pull metaphor
How do you build a framework?
(to build a common data infrastructure)
- Define goals
- Inventory existing initiatives
- Evaluate efficacy of existing initiatives
Structured process for student success insights
- Data is gathered from a variety of systems
- Data is normalized and modeled in a central repository
- Success metrics and risk scores are delivered via dashboard and CRM integration
Inventory and Segment Initiatives
- Student Lifecycle
- Near completer
- Student Attributes
- Student Supports
Analytics are used at critical lifecycle milestones to help them understand what is happening at the milestone and how to identify what services to offer for students at each stage.