Michigan Growth Model
Background reference for Adequate Growth Targets and Nearest Grade Average Method.
Most of what follows was on hidden slides 5–12 of TAC Growth Update Deck, April 2026 and was skipped in the live presentation. The minutes cover it because it came up in discussion.
Metrics
Student Growth Percentile (SGP) — the percentage of students with similar score histories having lower growth than this student. A normative description of growth achieved.
Adequate Growth Percentile (AGP) — the SGP a student must consistently meet or exceed to reach proficiency within a set timeframe. A criterion-referenced target expressed on a normative scale.
Growth target timeframe — the number of years in which a student is expected to reach proficiency. Michigan’s timeframes were set by empirical review of past student progressions.
Asked how the targets were originally set, the answer in the room was that they came out of “the R-set that Damian Betebenner built” — the R SGP package.
Coverage
- Content areas: grades 4–8 and 11. Grades 9 and 10 are computed and delivered to schools in the student data file but are not included in accountability.
- English language proficiency: grades 1–12.
Current target methods
| Method | Assessments | Precision |
|---|---|---|
| Quantile regression (R SGP package) | M-STEP, PSAT 8, PSAT 9, PSAT 10, SAT with Essay, WIDA ACCESS | Individual student score history — student, assessment, content area, grade, prior-year-1 and prior-year-2 scaled scores |
| Logistic regression | MI-Access | Group score history — assessment, content area, prior-year-1 performance level. A handful of large buckets |
The logistic regression build sets the growth target cut point at a 50% probability of proficiency.
The proposed Nearest Grade Average Method sits between the two on precision.
Timeframe rationale
- Content areas: 1–3 years. Three-year maximum on reliability concerns and Michigan legislative policy.
- English learner proficiency: 1–5 years. Five-year maximum on research (which suggests 5–7 years) and federal policy.
State-level uses
Public reporting — the Parent Dashboard, the Student Growth Report, the Annual Education Report. High-stakes school accountability through the School Index. The rationale is continuous improvement, and growth reporting is required under ESSA.
Two former uses no longer exist. From the deck’s speaker notes on slide 6, found nowhere in the minutes:
Educator Evaluations were repealed from state law. Districts may now conduct those however based solely on local policy and do not report them to the state. School grades were also repealed from state law.
This narrows the set of high-stakes consumers of growth data, and is relevant to any argument about how much precision the model must carry.