Student Growth: The Pulse of the Education System

DoWEA Fall Leadership Meeting · Executive Learning Session

Damian W. Betebenner

National Center for the Improvement of Educational Assessment

October 26–30, 2026

A brief introduction

Damian W. Betebenner, Ph.D.

National Center for the Improvement of Educational Assessment

Developer of the Student Growth Percentile methodology — the Colorado Growth Model — now used by education systems across the United States
Author of the SGP R package, the open implementation of that work
Adviser to states and large systems on using growth, not only achievement, for performance management

The views today are mine. They are offered as a conversation with DoWEA leadership, not as an audit of DoWEA schools.

How this session is organized

What student growth is about — a picture, not a formula
Death to proficiency. Long live proficiency. — why status alone is a poor management tool
Norm and criterion — how much did they grow, and is it enough?
DoWEA status and growth — NAEP achievement, and a first look at cohort growth
Growth for process management — finding pockets of strength and weakness
Questions

AI systems assisted in preparing this deck. Every number is tied to a cited public source or to labeled illustrative data. Responsibility for the claims remains with the author.

What student growth is about

Status: snapshots in time

Growth: learning between the snapshots

Same two snapshots. Advance. Now draw what happened between them: each line is one student, from where they stood in spring 2025 to where they stood in spring 2026. That flow is growth: the change in what students know and can do across time. Hold the two definitions side by side — status is a fixed point in time; growth is change across time.

Advance. Color each line by how that student’s progress compares with peers who started in the same place. Blue: more growth than most of those peers. Red: less. Grey: typical. That is a growth percentile, and we will come back to it.

Advance. This is the key contrast. Status systems compare Grade 4 this year with Grade 4 last year — the top row — which means comparing different groups of students. Growth follows the diagonal: the same students, followed over time. Both matter. They answer different questions.

Learning is a process.
Education is the system designed to support it.
Student growth is how we know whether that system is working.

Death to proficiency. Long live proficiency.

Status-based systems are poor performance-management tools

We live in a culture that loves ends and neglects means.

I want to be successful.
I want to win.
I want proficiency rates to go up.

Accountability systems built on a single status number ask schools to produce an outcome without describing the process that produces it.

The management error

Telling a school to “raise proficiency” is like telling a team to “win more.” It names the scoreboard. It does not name the work.

What sports teach about accountability

A coach is judged on wins and losses. That will not change.
Good coaches do not tell players to “win more.” They install a process.
If everyone plays their role, the team has the best chance of winning.
In education, wins and losses are proficient and not proficient.
“Raise proficiency rates” is the forced shot — an ill-advised play, written as policy.

The rational, maladaptive response

Focus on the bubble kids.

A status-only scoreboard makes it rational to spend the year on students just below the cut. That is the educational version of the forced shot. It can move the number. It is not a process for learning.

Wear two hats

Hat 1 — Accountability. Proficiency is what you are ultimately responsible for. A system where few students are proficient is not healthy, whatever else is true.

Hat 2 — Management. Effective performance management focuses on process. That is the hat you wear on Tuesday morning.

Nick Saban quote: know what you want to accomplish and focus on the process rather than the outcome.

Growth and status are opposite sides of the same coin

Where a student is today depends on where they started and how quickly they traveled.

Distance = Rate × Time

Growth is the leading indicator — the rate
Status is the lagging indicator — the distance

If you care about achievement, you already care about growth. You may not be watching it yet.

Infographic ordering processes, impacts, and outcomes, with growth as the leading indicator and status as the lagging indicator.

Norm and criterion

Two ways of understanding the world

Stylized pediatric height chart with percentile curves and a red criterion line labeled tall enough to ride.

An inch is not always an inch

Judy the elite jumper gains two inches; Anna the novice gains six inches.

A complete understanding requires both

Norm-referenced. Where does this observation sit in a distribution of similar observations?

Criterion-referenced. Where does this observation sit relative to a standard we care about?

For student growth

How much did the student grow? — among academic peers.

Is it enough? — relative to a proficient (or college-ready, or on-track) standard.

Student Growth Percentiles: how much did they grow?

Percentile growth trajectories: is it enough?

Percentile growth trajectories from one Grade 3 starting score below Proficient. Typical growth never reaches the rising Proficient line by Grade 8; growth near the 60th percentile every year reaches it by Grade 8, and higher percentiles reach it sooner.

Growth and system status

Conceptual four-quadrant scatter of status versus growth.

DoWEA status and growth

DoWEA students are high achieving

Dumbbell chart of 2024 NAEP average scale scores, DoWEA versus national public, for mathematics and reading in grades 4 and 8.

The gap is not only the average

Bar chart of percent at or above NAEP Proficient in 2024, DoWEA versus national public: Grade 4 math 54 versus 39, Grade 8 math 41 versus 27, Grade 4 reading 48 versus 30, Grade 8 reading 53 versus 29.

Steady through the pandemic years

Four panels, 2017 to 2024, DoWEA versus national public NAEP averages. DoWEA is flat in every panel; national public declines in every panel.

DoWEA held steady from 2017 to 2024 while national averages fell. That is a status success. It still does not answer whether students are learning a lot.

Our students are high achieving.
But are they learning a lot?

How to read a growth number from NAEP

NAEP does not follow the same students from Grade 4 to Grade 8. It takes two snapshots of different samples from the same cohort — here, Grade 4 in 2015 and Grade 8 in 2019, mathematics.

Recent work with Professor Henry Braun estimates cohort growth from those two snapshots. It borrows the link between Grade 4 and Grade 8 scores from states that do follow individual students.

The result is a mean SGPc: the average growth percentile of a jurisdiction’s students against national academic peers.

A mean of 50

The jurisdiction’s students progressed, on average, as their national peers did from the same Grade 4 starting points.

DoWEA: 54.1 — seventh of 52 jurisdictions, in the very high group. On average, DoWEA students finished Grade 8 at the 54th percentile of national peers who started Grade 4 in the same place. (NAEP reports DoWEA as DoDEA.)

DoWEA growth among the highest in the nation

Dot plot of mean SGPc for 52 jurisdictions, Grade 4 2015 to Grade 8 2019 mathematics. Only DoWEA is labeled; it sits in the leading group.

High status and high growth

Scatter of 2015 Grade 4 NAEP mathematics status against mean SGPc for unlabeled jurisdictions, with DoWEA labeled in the high-status high-growth quadrant.

One level up: the United States internationally

Dot plot of mean SGPc for 33 education systems in TIMSS, Grade 4 2015 to Grade 8 2019 mathematics. Only the United States is labeled; it sits in the typical group near the international middle.

Same question, same cohort years, one level up: among education systems, U.S. growth is typical, near the international middle.

Same method, same cohort design — Grade 4 in 2015 to Grade 8 in 2019, mathematics — but now on TIMSS, and the peers are students in other education systems. Most are countries; a few are provinces or regions that TIMSS reports separately. Other names are deliberately unlabeled.

The United States sits at 47.4, with a sampling interval of 43.5 to 51.3 that includes 50. That is typical growth: U.S. students progressed about as their international academic peers did. By point estimate it is 23rd of 33, but rank is not the point.

Expect the question “so DoWEA is above a nation that is below the world?” Do not chain the numbers. NAEP compares DoWEA with U.S. peers; TIMSS compares the United States with international peers. Different tests, different peers. Each answers the same question at its own level.

Comparisons require a common test

Each step up trades local fit for wider comparison.

International assessment

TIMSS. Compares the United States with other countries.

National assessment

NAEP. Compares every participating jurisdiction — DoWEA, each state, and the nation.

Consortium assessment

Smarter Balanced and similar. Compares the members that share the test.

Unique assessment

DoWEA CCRS summative (grades 3–7). Compares DoWEA schools and students with one another.

DoWEA’s CCRS summative assessments are the right instrument inside the system. NAEP is the bridge across states; TIMSS is the bridge across countries.

Growth for process management