Dashboard Insights FAQ
Following are the answers to most frequently asked questions about the data insights that appear on the advanced Dashboard of Advising.
Does every unit see the same dashboard?
Not necessarily. Advising users see the dashboard as it’s been set up for their own Student Success Unit (SSU). If that SSU doesn’t work with data insights, their dashboard may have them hidden. Certain sensitive fields may have been hidden as well, as a customization for your institution.
Persistence Predictions
Doesn’t a persistence score treat students as a number?
Not if you understand the prediction as being a data-driven tool for early warning of who’s at risk. A student’s “score” is one signal among many, but one of the ones that let you get ahead of problems:
- Reactive signals: prior term GPA, academic standing
- Proactive signals: alerts raised, LMS login relative to their peers
The persistence score surfaces at-risk students you might otherwise miss because their reactive signals, such as GPA, seem fine.
The model is trained on the most recent years of student data, and the scores help you track changes in behavior: how a student may be disengaging (dropping in persistence score) or re-engaging (increasing their persistence score). This lets you prioritize groups of students and allocate the right resources at the right times during a term to help the most. The score gives you insight to bear on the many challenges and policies you’re already juggling, such as a student’s perceived fit with their major and institution, ideal course load and course modality, financial aid, or what they might divulge about their mental or emotional wellness or outside commitments.
How is a persistence score not deterministic?
That’s the opposite of its purpose. It’s not a score (evaluation) but a signal (prediction), a helpful warning of where things are headed if you don’t intervene. As Scrooge wonders when facing his tombstone, “Why show me this, unless it can change?”
Ultimately, the persistence score is a shortcut for you as an advisor, to know where your time is best spent. When you have too much to do, the signal guides you to choose well among all the students you could engage, to maximize your impact.
How can a strong student have low persistence predicted?
This unintuitive pairing might be pointing to a crisis in the student’s situation: their long-term indicators match well for academic success, but some behavioral change triggered the model to predict a near-term problem staying enrolled. Pay attention to these flags: you might uncover a family, financial, or health crisis needing timely support.
For example, if a strong student were having a serious family crisis, the prediction might be reflecting behavioral changes in, say, delayed/blocked registration, a steep drop in in-progress grades, or receiving academic alerts.
Engagement Scores
If a unit won’t use Advising, does that affect students’ engagement scores?
Yes, having an area fail to use Advising for scheduling and communication does disrupt the usefulness of tracking “engagement”. A student’s Advising “engagement” is counted as any note, email, or SMS message created for them within the last 30 days (which is a dynamic, rolling count). If a unit schedules meetings outside of the application, for example, that activity cannot be rolled up into the full picture of that student’s advising experience.
Urge your non-compliant units to make full use of Advising for scheduling and communication. Let them know that operating outside of Advising forfeits both the engagement measures and all historical management of their Advising activity.
Is “low course engagement” per course? What if the instructor doesn’t use the LMS?
It’s better than by course: it’s compared by section. By comparing to the other students in the same section, it corrects for situations such as one of the instructors bypassing the LMS, which lowers usage.
Do those who withdraw count against engagement stats for that course?
No; that’s corrected for. Once a student withdraws, their engagement is not reflected in the LMS engagement tile.