NM-MSSA Assessment Data & Analysis
This dashboard brings together tools for understanding New Mexico student proficiency on the NM-MSSA. Choose a section below to get started.
This interactive tutorial walks you through the key concepts needed to interpret NM-MSSA student growth trajectory charts. Understanding these charts is essential for correctly reading the proficiency data in this dashboard.
| Subgroup | Slope (pts/yr) | Projected Year |
|---|---|---|
| Statewide | +0.61 | Not in K-12 |
| Econ. Disadvantaged | +0.58 | Not in K-12 |
| English Learners | +0.80 | Not in K-12 |
| Students w/ Disabilities | +1.80 | Not in K-12 |
| Native American | +0.95 | Not in K-12 |
| Subgroup | Slope (pts/yr) | Projected Year |
|---|---|---|
| Statewide | +1.04 | Not in K-12 |
| Econ. Disadvantaged | +1.12 | Not in K-12 |
| English Learners | +1.50 | Not in K-12 |
| Students w/ Disabilities | +1.40 | Not in K-12 |
| Native American | +1.90 | Not in K-12 |
How the growth projections and proficiency transition visualizations in this dashboard were generated.
All analyses use student-level results from the New Mexico Measures of Student Success and Achievement (NM-MSSA) for school years 2021–22 through 2024–25. Students were organized into longitudinal cohorts defined by their grade level and test year, allowing us to follow the same students across consecutive grades. Reading/language arts and math were analyzed separately throughout.
The dashboard features the Grade 4 cohort of 2022—students who were in 4th grade in SY22 and had reached 7th grade by SY25. This cohort was selected because it had four complete years of NM-MSSA scores and fell in the middle of the grade range over which the test is administered. The cohort includes 26,053 students in reading and 26,126 in math.
All projections on the Proficiency Trajectories tab—statewide and subgroup alike—are generated using the same linear mixed-effects (LME) model, estimated in R using the lme4 package. This unified approach ensures that every group’s trajectory is modeled with the same statistical rigor, accounting for both group-level trends and individual student variation.
The model treats each student’s test scores as repeated observations over time. It estimates two things: an average starting score in 4th grade (the intercept) and an average change in score per year (the slope). Time is measured in grades and centered at Grade 4.
The model allows each student to have their own starting score and growth rate (random effects), capturing the wide variation in individual trajectories around the group average. The projected lines on the Proficiency Trajectories tab extend each group’s estimated average growth rate forward from the model intercept.
The table below shows the estimated average growth rate (slope) for each student group, along with the standard error of that estimate. The “Projected Year” column indicates when the group is projected to reach the proficiency benchmark of 60, based on its current trajectory. Groups that are not projected to reach proficiency before the cohort’s expected high school graduation in 2031 are marked “Not in K-12.”
| Subgroup | Mathematics | Reading | ||
|---|---|---|---|---|
| Slope (SE) | Proj. Year | Slope (SE) | Proj. Year | |
| Statewide | +0.61 (0.04) | Not in K-12 | +1.04 (0.04) | Not in K-12 |
| Econ. Disadvantaged | +0.58 (0.06) | Not in K-12 | +1.12 (0.06) | Not in K-12 |
| English Learners | +0.80 (0.09) | Not in K-12 | +1.50 (0.09) | Not in K-12 |
| Students w/ Disabilities | +1.80 (0.10) | Not in K-12 | +1.40 (0.11) | Not in K-12 |
| Native American | +0.95 (0.13) | Not in K-12 | +1.90 (0.13) | Not in K-12 |
Time to proficiency is calculated as:
No student group—statewide or subgroup—is projected to reach the proficiency benchmark before the cohort graduates high school in 2031 in either subject. In math, even the fastest-growing group (Students w/ Disabilities at +1.80 pts/yr) would not reach 60 until approximately 2038. In reading, the closest groups (Statewide and Native American) fall just short, reaching proficiency around 2032.
The growth rate for each subgroup is an average—a single number summarizing thousands of individual student trajectories. But how confident should we be in that average, and how much do individual students vary around it? The charts below answer both questions.
Each chart shows two layers of uncertainty for every subgroup:
Dark bars (Average Growth 95% CI): This is the confidence interval around the group’s estimated average growth rate. It tells you how precisely the model has estimated the group average. A narrow interval means the model is quite confident in the average; a wider one means less precision. If we repeated this study with a new sample of students, we’d expect the true average to fall within this range 95% of the time.
Light bars (Individual Student 95% Range): This is the spread of individual student growth rates—where roughly 95% of students in the group actually fall. This range is always much wider than the confidence interval, because students vary enormously. Some grow several points per year; others decline. The projection line on the Proficiency Trajectories tab represents the group average, not a guaranteed path for any one student.
The Proficiency Transitions tab shows how individual students moved between proficiency categories from SY24 to SY25. Each student is classified into one of four groups: Stayed Proficient, Improved to Proficient, Declined to Not Proficient, or Stayed Not Proficient. The treemap visualizations display these transitions aggregated across grades 4–8, broken out by district/charter, ELL status, disability status, and free/reduced lunch eligibility.
Schools with fewer than 10 total students in the data are masked entirely to protect student privacy, and individual cell values below 10 are suppressed in tooltips.
Additional caveats to keep in mind:
Scale score comparability. The NM-MSSA technical report does not establish a vertical scale across grades, meaning a one-point gain may not represent the same amount of learning in 4th grade as in 7th grade. Growth rates estimated across grades should be treated as approximations.
Measurement error. Scale scores carry a margin of error of 2–3 points, and a student’s performance level can shift from year to year due to factors beyond academic growth.
Four observed years. All models are fitted to just four years of NM-MSSA data (SY22–SY25). While the mixed-effects approach uses individual student trajectories, the projections still extrapolate from a short observation window and should be treated as trend indicators rather than forecasts.
Triangulation. NM-MSSA results should be considered alongside other evidence of student learning, not used in isolation.