According to the National Center for Education Statistics (NCES), average reading and mathematics scores for 9-year-olds increased in 2025 compared with 2022, while 13-year-olds showed no significant change in either subject compared with 2023. This explainer covers what the long-term trend (LTT) assessment is, how it differs from the better-known main NAEP, and how to read its numbers without overstating them.
The results come from the National Assessment of Educational Progress, often called the Nation’s Report Card. As of Nov. 11, 2026, the NCES long-term trend page shows a last-updated date of May 28, 2026.
What is the Nation’s Report Card?
NAEP measures student achievement across the United States and, according to NCES, has informed education policy and practice since 1969. Congress requires it. NCES, part of the U.S. Department of Education’s Institute of Education Sciences, carries out the work, while the National Assessment Governing Board, which NCES describes as an independent body of educators, community leaders and assessment experts, sets NAEP policy.
The assessments are given to a sample of students in public and private schools, not to every student. NCES selects schools so that the participating students represent their district, state or the nation. That design is why NAEP is a tool for judging systems, not individual children or individual schools.
According to NCES, the main program covers these subjects:
| Subject | Grades assessed | Reporting level |
|---|---|---|
| Reading | 4, 8, 12 | National, state, district |
| Mathematics | 4, 8, 12 | National, state, district |
| Science | 4, 8, 12 | National; state for grades 4 and 8 |
| Writing | 8, 12 | National |
| Arts | 8 | National |
| Economics | 12 | National |
NCES also lists civics, geography, U.S. history and technology and engineering literacy. State availability for some subjects varies by year.
What is the long-term trend assessment?
The long-term trend assessment is a separate strand of NAEP. According to NCES, since the 1970s it has measured student performance in mathematics and reading through LTT assessments. Reading began in 1971 and mathematics in 1973. The title of the NCES page frames the record as “1970-2025 Trends.”
The purpose is stability. Because the test content and methods are kept as consistent as possible, scores can be compared across decades in a way that is harder when a test is revised often. NCES notes a break in the series around a 2004 redesign, which is one reason comparisons with older results call for caution.
LTT students are selected by age, not grade. The assessment covers ages 9, 13 and 17. The 2025 round included about 15,000 students at age 9 and about 16,400 at age 13, according to NCES, so the 2025 results cover those two ages.
How does the long-term trend test differ from main NAEP?
The two programs are easy to confuse, and NCES says they cannot be compared directly. The differences, as NCES lays them out:
| Feature | Long-term trend | Main NAEP |
|---|---|---|
| Who is sampled | Ages 9, 13 and 17 | Grades 4, 8 and 12 |
| Reading began | 1971 | 1992 |
| Math began | 1973 | 1990 |
| Frequency | Every four years since 2004 | Every two years, in odd-numbered years |
| Scale | 0 to 500 | 0 to 300 or 0 to 500, depending on subject |
| Result labels | Performance levels from 150 to 350 | Basic, Proficient and Advanced achievement levels |
Content differs as well. According to NCES, the long-term trend math test emphasizes numbers and numeration, variables and relationships, shape, size and position, measurement, and probability and statistics, along with basic skills and definitions. Main NAEP math also covers algebra and geometry and puts more weight on problem solving and reasoning. In reading, the long-term trend test uses short narrative, expository and document passages and asks students to locate information, make inferences and find the main idea. Main NAEP uses a wider range of texts, including poetry and procedural texts, and asks for more complex inferences and comparisons across texts.
One practical consequence: a headline about “NAEP scores” can refer to either program. A reader should check whether a figure is for fourth graders, which points to main NAEP, or 9-year-olds, which points to the long-term trend.
What did the 2025 results show?
Two findings stand out, according to NCES.
- Age 9: average scores in reading and mathematics increased in 2025 compared with 2022.
- Age 13: there were no significant changes in reading or mathematics compared with 2023.
The comparison years differ because the two ages were last assessed in different years. The linked NCES publications cover a 2022 highlights report for age 9 and a 2023 highlights report for age 13.
The NCES page reviewed for this article does not publish the score points or percentile breakdowns in text, so this article does not quote them. Readers who want the exact averages, subgroup results or changes at lower and higher percentiles can consult the full NAEP report for 2025 through NCES.
Why does “no significant change” not mean “no change”?
The phrase has a precise meaning. NCES says that differences are discussed only when they are statistically significant at the 0.05 level, with adjustments for multiple comparisons. In plain terms, a result is called significant when a gap that size would be unlikely to appear by chance if nothing had really changed among all students.
That leads to a point NCES stresses: “significant” refers to statistical reliability, not to the size or educational importance of a difference. A small gain can be significant if the sample is large and the estimate precise. A larger-looking move can fail the test if the uncertainty is wide.
For 13-year-olds, then, the accurate reading is that the 2025 sample did not show a change that could be distinguished from sampling noise. It is not evidence that scores rose or fell.
Estimates, not exact values
NCES reminds users that results come from representative samples, so they carry sampling error. Standard errors, which show that uncertainty, are available through the NAEP Data Explorer. Any average score, in other words, is best thought of as a range around a central estimate.
How should the scale and performance levels be read?
Average scores are reported on a 0 to 500 scale for both reading and mathematics. NCES says each subject’s scale was derived independently, so reading and math averages cannot be compared with each other. A 9-year-old average in reading cannot be said to be “higher” than the math average.
The results also show the percentage of students at or above five performance levels, which correspond to scale points 150, 200, 250, 300 and 350. Each level is tied to more advanced skills and knowledge. The long-term trend uses these descriptive levels instead of the Basic, Proficient and Advanced labels of main NAEP, so a percentage “at or above 250” is not the same as a percentage “proficient.”
A worked example of the thresholds
The five levels are evenly spaced 50 points apart, from 150 to 350. As illustrative arithmetic only: the span from the first level to the fifth is 350 minus 150, or 200 points, which is four steps of 50. A score gain of 5 points, which is a tenth of one step, would be small against that ladder even if it were statistically significant. The example is hypothetical and does not describe any 2025 result.
What cautions does NCES attach to the numbers?
NCES lists several limits that apply to any use of LTT data.
- Trend breaks. The 2004 redesign and later changes to accommodations and procedures mean comparisons with 2004 or earlier results need caution.
- Not comparable to main NAEP. The instruments and methods differ.
- Subgroup thresholds. Subgroups are reported only when they meet minimum sample and school requirements, though all students count toward overall results.
- Age-based samples. Because students are sampled by age rather than grade, grade-level comparisons need care.
- No causal claims. NCES cautions users “against interpreting NAEP results as implying causal relationships.” Differences between groups, or links to teaching practices, should be read alongside socioeconomic and educational factors.
The last point matters most when readers are tempted to credit or blame a particular policy for a score change. A national sample of 9-year-olds can show that scores moved. It cannot show why.
How does this connect to Maryland classrooms?
The long-term trend assessment reports national results for these ages, so it does not rank Maryland or any other state. For the state picture, the main NAEP, which reports at the state level for reading and mathematics in grades 4 and 8, is the relevant tool, along with state assessments and local data.
Readers can find context on the system behind those results in the Maryland Herald explainer on how Maryland public schools are organized into 24 school systems. Attendance is a related factor in achievement, which is covered in what chronic absenteeism means and what helps schools. For the national enrollment backdrop, see public school enrollment and what federal data shows.
Reading the next education headline
A few habits make test-score coverage easier to judge.
- Identify the program: long-term trend by age, or main NAEP by grade.
- Check the comparison years, since the two programs are on different schedules.
- Look for the word “significant” and remember what it does and does not mean.
- Treat averages as estimates with a margin of error.
- Be wary of any claim that a single policy explains a national score change.
Taken together, the 2025 long-term trend results show a measured improvement for 9-year-olds and a flat picture for 13-year-olds, reported with the caveats NCES attaches to every NAEP release. The long record that began in the early 1970s is the main reason the test remains useful, and the reason its limits are worth understanding before drawing conclusions.
This article is general information about education statistics and is not individual advice for any family or school.
Sources: National Center for Education Statistics
