How SchoolZone.ai Rates Schools
We measure what schools actually contribute to student learning — not the wealth of the neighborhoods they serve.
The Problem with Traditional School Ratings
Academic research consistently shows that 60–70% of the variance in standardized test scores is explained by socioeconomic status, not school quality. Traditional rating systems like GreatSchools primarily reflect neighborhood wealth: affluent communities with college-educated parents score high, while under-resourced communities score low — regardless of whether their schools are actually doing a good job.
This creates a feedback loop: families flock to “high-rated” schools, driving up housing prices, which further concentrates wealth. Meanwhile, schools doing remarkable work with fewer resources are labeled as “failing” simply because their students start from a different baseline.
Our Approach: Value-Added Methodology
The SchoolZone Effectiveness Score uses a value-added methodology that asks a fundamentally different question: “Given the students this school serves, are outcomes better or worse than expected?”
Instead of simply ranking raw test scores, we use regression analysis to control for the percentage of students eligible for free or reduced-price lunch (a standard proxy for economic disadvantage). Schools that outperform their expected outcomes are rated higher, regardless of their absolute test scores.
How It Works
Group Schools
Schools are grouped by state and level (elementary, middle, high) so comparisons are fair — an elementary school in Texas is compared to other Texas elementary schools, not high schools in New York.
Fit the Expectation Line
Within each group, we use linear regression to find the expected relationship between economic disadvantage (% free/reduced lunch) and academic proficiency. This tells us what proficiency level we'd expect given a school's demographics.
Calculate the Residual
For each school, we calculate the residual: actual proficiency minus expected proficiency. A positive residual means the school is doing better than expected; a negative residual means worse.
Assign the Score
Schools are ranked by their residual within each state/level group and placed into deciles (1–10). A school at decile 9 is outperforming 80–90% of similar schools.
Three Rating Tiers
Outperforming
Scores 8–10. These schools achieve significantly better academic outcomes than expected given their student demographics. They demonstrate exceptional educational effectiveness and add substantial value.
On Track
Scores 4–7. These schools produce outcomes roughly in line with what's expected for their student population. Students are making typical progress.
Needs Support
Scores 1–3. These schools are achieving below expected outcomes and may benefit from additional resources, staffing, or programmatic support. This rating reflects opportunity for improvement, not a judgment of the community.
Data Sources
NCES Common Core of Data (CCD)
The U.S. Department of Education's annual census of all public schools and school districts. Provides enrollment, free/reduced lunch eligibility, and school characteristics. We use the most recent available data.
nces.ed.gov/ccdStanford Education Data Archive (SEDA) 6.0
Produced by the Stanford Graduate School of Education, SEDA provides standardized test score data linked to the NAEP scale for grades 3–8, enabling fair comparisons across states with different testing standards. SEDA covers approximately 2009–2018 school years.
edopportunity.orgLimitations & Transparency
- SEDA data vintage: The test score data comes from approximately 2009–2018. While this is older, value-added ratings measure relative performance patterns, which tend to be stable over time.
- Grade 3–8 only: SEDA covers grades 3–8 (state testing grades). High schools don't have fair ratings in version 1. Future versions will incorporate graduation rates as an outcome variable.
- Free/reduced lunch as proxy: FRL eligibility is a coarse proxy for socioeconomic status. It doesn't capture the full spectrum of advantage and disadvantage, and some states have adopted community eligibility provisions that make all students eligible.
- Minimum group size: State/level combinations with fewer than 30 schools don't receive ratings, as regression estimates would be unreliable.
- Private schools: Private schools don't participate in state testing and aren't included in CCD data. They are marked as “Not Rated.”
- Not a single measure of quality: No rating captures everything about a school. We encourage families to visit schools, talk to teachers and parents, and consider factors beyond test scores.
Frequently Asked Questions
Why not just use GreatSchools or Niche ratings?
We display GreatSchools and Niche ratings where available — they provide useful information. But their methodologies heavily weight raw test scores, which correlate strongly with family income. Our fair rating adds a complementary perspective: which schools are doing the most with what they have.
Can a school in a wealthy area receive a low fair rating?
Yes. If a school in an affluent area has high test scores but those scores are lower than expected given the student population's advantages, it will receive a lower fair rating. The school may still have good absolute outcomes, but it isn't adding as much value as comparable schools.
Can a Title I school receive an “Outperforming” rating?
Absolutely — and many do. That's the whole point. Schools serving high-poverty populations that achieve better outcomes than expected are recognized for the real educational value they provide. These are the hidden gems that traditional ratings often miss.
Why is my school “Not Rated”?
A school may not have a fair rating if: (1) it's a private school, (2) it doesn't report free/reduced lunch data, (3) SEDA test score data isn't available for it, or (4) there are fewer than 30 schools in its state/level comparison group.