Retention And Student Success

---
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:

1. **On the edge + saveable** → highest-leverage first touch now.
2. **At risk, longer runway** → structured outreach this term.
3. **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

1. Attach the definition to every retention/persistence number (`[verify-at-use]`).
2. Segment the loss (or the at-risk cohort) by cause and term.
3. Triage by leverage; place the first touch where it changes an outcome.
4. Read the gateway DFW courses feeding the loss; fix the course, not just the student.
5. 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

- Traverse the **at-risk student triage** and **enrollment-vs-retention lever choice** trees in [`../../knowledge/higher-ed-decision-trees.md`](../../knowledge/higher-ed-decision-trees.md).
- [`../registrar-and-academic-operations/SKILL.md`](../registrar-and-academic-operations/SKILL.md), [`../../templates/retention-intervention-plan.md`](../../templates/retention-intervention-plan.md).
- Best practices: [`../../best-practices/retention-is-an-early-alert-problem.md`](../../best-practices/retention-is-an-early-alert-problem.md).