| name | retention-and-student-success |
| description | Turn retention into an early-alert operation: compute persistence with its definition attached, triage at-risk cohorts by leverage, and read DFW gateway courses as retention infrastructure. Metric definitions vary by IPEDS/accreditor/IR -> verify-at-use; cohort-level only, no student PII. |
Retention & Student Success
Retention is a lagging number produced by leading indicators. This skill works the leading edge — early-alert signals, at-risk triage, and gateway-course performance — so the intervention lands while the window is still open.
Advisory, not academic-policy or clinical advice. Retention and persistence metrics are defined differently by IPEDS, the accreditor, and IR — every one carries its definition + retrieval date + [verify-at-use]. Work in cohorts and risk segments, never a named student's record — individual data goes back to the early-alert/SIS system. No PII.
Define before you measure
| Term | Common variants (attach yours) | Note |
|---|
| Retention | Fall-to-fall vs term-to-term; first-time-full-time cohort vs all-entering | A rate is uncomparable without this |
| Persistence | Continued enrollment anywhere vs at this institution | IPEDS vs institutional differ |
| Completion / graduation | 100% / 150% / 200% of normal time; which cohort | Accreditor and IPEDS definitions vary |
| DFW rate | Share of a course's students earning D / F / Withdraw | The gateway-course health signal |
The rule: state cohort, term, and source with every number. [verify-at-use].
Early-alert leading indicators
- First-weeks attendance and LMS engagement.
- First-gradebook signals in gateway courses.
- Unmet financial need / aid gaps at census.
- Advising no-shows and registration holds.
At-risk triage by leverage
Group the cohort — don't treat every flag equally:
- On the edge + saveable → highest-leverage first touch now.
- At risk, longer runway → structured outreach this term.
- Flag without cause → monitor, don't spend the touch.
Metrics table
| Metric | What it tells you | Watch for |
|---|
| Fall-to-fall retention | Year-over-year persistence | Reported without cohort/term definition |
| Term-to-term persistence | The nearer-term signal | The place a fall-to-fall drop was already visible |
| DFW rate (gateway) | Course-level retention risk | A gateway feeding many majors |
| Early-alert response rate | Whether the operation is working | Alerts raised but not worked |
Workflow
- Attach the definition to every retention/persistence number (
[verify-at-use]).
- Segment the loss (or the at-risk cohort) by cause and term.
- Triage by leverage; place the first touch where it changes an outcome.
- Read the gateway DFW courses feeding the loss; fix the course, not just the student.
- Flag any enrollment-quality signal back to enrollment management.
Anti-patterns
- Quoting retention with no definition or source.
- Intervening at the withdrawal desk instead of the first weeks.
- Treating a high-DFW gateway course as a student problem, not an infrastructure one.
See also
