Student Success by DesignA practitioner’s guide to theory, data, and AI in student success

Section 7

Student Success and Data Governance

Who decides how student data is defined, shared, and used, and how those decisions protect students while making success work possible.

In this section

7.1Why data governance is a student success issueEvery early alert, dashboard, and AI assistant depends on decisions about data that someone made, or failed to make.Key takeaway. Governance is part of student success work.7.2Governing the student data lifecycleGovernance decisions apply at every stage, from collection to deletion.Key takeaway. Most governance failures happen after launch, when data is reused, combined, or kept longer than anyone intended.7.3Roles, decision rights, and student voiceEffective governance is cross-functional. Student success data crosses too many offices for any one of them to govern alone.Key takeaway. If students are absent from governance, the institution is guessing about what students consider acceptable.7.4Laws and policies that applyCompliance is the minimum. Student success governance also asks what is appropriate.Key takeaway. “Legal” and “trustworthy” are different tests. Student success work needs to pass both.7.5Data quality and shared definitionsA model is only as fair and accurate as the records it is built on.Key takeaway. The students least visible in the data are often the ones support is meant to reach.7.6Getting started: a governance checklistUse this list to assess an existing program or to launch one alongside a new student success tool.Key takeaway. Start with one tool and one lifecycle map. Governance grows from use cases, not from policy documents alone.7.7Decision rights and impact assessmentTwo templates: who decides what about student data and algorithms, and what to assess before any model affects students.Key takeaway. Governance tests the institution’s capacity to act on data responsibly. Model accuracy is one input.7.8Monitoring, escalation, and retirementApproval is the start. These templates cover what to watch after launch, what to do when something goes wrong, and when to stop.Key takeaway. Set the retirement criteria at approval, while the decision is still neutral.