Episode 83

From LMS Signals to Student Success: Learning Analytics, Data Governance & AI Ethics

Guest name: Thomas Cavanagh
Guest title: Vice Provost for Digital Learning
Guest company: University of Central Florida
Host: David Brunner

Brief description of the episode

What does a cultural shift have to do with learning analytics? Thomas Cavanagh, Vice Provost for Digital Learning at the University of Central Florida, explains how higher ed institutions can move data to action by changing the mindset around data governance.

This episode uncovers important dynamics between shared data governance, key LMS metrics, and student retention. The conversation dives deep into AI for higher education, predictive models, data ethics, and human oversight when using analytics to inform institutional strategy.

Key Takeaways:

- Institutions must first build a culture that is curious about data and willing to use it for decision-making.
- Then, just collecting data alone is not enough; it must lead to concrete actions, such as student interventions or changes to advising.
- Analytics should surface actionable insights, such as identifying course combinations that negatively affect GPA.
- Data cannot remain in silos across departments; coordination and shared governance are essential.
- A centralized role or structure with a “bird’s-eye view” helps ensure data is accurate, connected, and leveraged institution-wide.
- Adoption numbers only tell part of the story and are not the most meaningful indicator of success. Institutions should prioritize metrics tied to student outcomes, such as retention, engagement, grade improvement, and speed to graduation.
- Learning management system data is especially valuable because it provides near real-time indicators of student performance.
- Early signals, such as test scores, registration behavior, and course activity, enable proactive support before students fall off track.
- Analytics investments should be evaluated based on measurable institutional impact, not tool usage.
- Universities should actively teach students to use AI ethically, as it will be expected in many careers.
- AI is well-positioned to support course development, student support, and scalable teaching assistance.
- A 24/7 AI teaching assistant inside courses could handle common student questions and support large class sizes.
- Human oversight must remain central, especially in assessment, grading, and high-stakes academic decisions.
- Over-reliance on AI should be avoided, as models can reinforce bias or tell users what they want to hear if not validated.
- Institutions should be transparent about what data is collected, how it is used, and what users can opt out of.
- Ethical data handling includes protecting sensitive student information and preventing improper use in open AI systems.
- Predictive and analytics models should be tested against historical data to validate accuracy before deployment.
- While data minimization is important, uncollected data cannot be analyzed later, creating a strategic tradeoff.
- Faculty commitment to student success, affordability, and ethical technology adoption remains the strongest driver of fair and effective digital learning decisions.

From LMS Signals to Student Success: Learning Analytics, Data Governance & AI Ethics

Brief description of the episode

Key Takeaways:

You may also like this

Get the Citation‑First AI Checklist