About NYC School Explorer
What This Tool Does
NYC School Explorer is an AI-powered tool for exploring NYC School Quality Report data. It helps journalists, researchers, and educators understand patterns in school performance with appropriate context and limitations.
Unlike simple ranking sites, this tool emphasizes student growth (Impact Score) over absolute test scores (Performance Score), and always presents metrics alongside poverty context (Economic Need Index).
Data Sources
- NYC DOE School Quality Reports (2022-23, 2023-24, 2024-25)
Elementary/Middle (EMS), High School (HS), Transfer (HST), Early Childhood (EC), D75. Impact Score and Performance Score available for 2023-24 and 2024-25 only. ENI, survey results, attendance, and ratings available for all three years. - School Budgets (Local Law 16) (2022-23, 2023-24, 2024-25)
Fair Student Funding allocations, total budget, % funded, and gap to full funding. Note: Charter school budgets are not directly comparable to DOE-managed schools. - Student Discipline (Local Law 93) (2022-23, 2023-24, 2024-25)
Removals, principal suspensions, and superintendent suspensions by school. Small counts (1-5) are redacted for student privacy. - PTA Financial Reporting (2022-23, 2023-24, 2024-25)
Income, expenses, and ending balances by school. PTA income reflects parent wealth, not school quality. - School Locations (LCGMS)
Addresses, coordinates, building codes, neighborhood (NTA), council district, and principal information.
Key Metrics Explained
Impact Score (Student Growth)
Measures how much students learn relative to similar students citywide. A score of 0.60 means students grew more than 60% of similar students.
Correlation with poverty: r = -0.29 (weaker than Performance Score)
Performance Score (Absolute Outcomes)
Measures absolute test scores and outcomes. Reflects where students are, not how much they grew.
Correlation with poverty: r = -0.69 (strong negative correlation)
Economic Need Index (ENI)
Poverty indicator based on temp housing status, HRA eligibility, and free lunch data. Higher values indicate higher poverty.
Citywide median: 0.87 for 2024-25 (varies by school type and year)
School Categories
Note on Thresholds: Categories are computed for all school types using the same thresholds (Impact ≥ 0.55, Performance ≥ 0.50). These thresholds were validated using Elementary/Middle School (EMS) data, where they represent approximately the 75th percentile. The same percentile cutoffs may differ for High Schools and other school types.
High-poverty schools (ENI ≥ 0.85) are classified based on their Impact and Performance scores:
| Category | Impact | Performance | Meaning |
|---|---|---|---|
| Strong Growth + Strong Outcomes | ≥ 0.55 | ≥ 0.50 | Strong growth AND strong absolute scores |
| Strong Growth, Building Outcomes | ≥ 0.55 | < 0.50 | Strong growth despite lower absolute scores |
| Strong Outcomes, Moderate Growth | < 0.55 | ≥ 0.50 | Strong scores but moderate growth (rare) |
| Developing on Both Metrics | < 0.55 | < 0.50 | Neither strong growth nor strong scores |
Schools with ENI below 0.85 are classified as "Lower Economic Need" and excluded from this framework.
Persistent High Growth
Schools that maintained "Strong Growth" status (Impact ≥ 0.55) across both 2023-24 and 2024-25. Two consecutive years of high growth suggests more consistent patterns, though we still cannot determine why these schools produce strong results.
Why the High Growth Page Focuses on EMS
While categories are computed for all school types, the High Growth Schools page defaults to Elementary/Middle Schools (EMS). This focus exists for several reasons:
- Threshold validation: The thresholds (Impact ≥ 0.55, Performance ≥ 0.50) were calibrated using EMS data where they represent approximately the 75th percentile. The same thresholds may represent different percentiles for High Schools.
- Different metrics: High School reports include graduation rates and other metrics not present in EMS reports, which may affect how Impact Score is calculated.
- Sample size: EMS represents the largest group of schools, providing more robust patterns.
- Methodological consistency: Focusing on one school type ensures findings are internally consistent and comparable.
You can explore High Schools and other school types through the search and chat features. Categories are available for all school types, but interpret with caution since the thresholds were not specifically validated for non-EMS schools.
Important Limitations
- •Only 2 years of data: Impact Score data is available for 2023-24 and 2024-25 only. This limits our ability to identify long-term patterns.
- •Methodology not fully disclosed: The NYC DOE has not published the complete methodology for calculating Impact Score.
- •Cannot prove causation: Correlations and patterns do not prove what causes high or low performance. Multiple explanations exist for every finding.
- •No student mobility data: We cannot rule out selection effects (students leaving or being counseled out).
- •Year-over-year volatility: Many schools change category between years. A single year's data may not reflect long-term patterns.
Responsible AI Design
This tool includes multiple guardrails to prevent harmful use:
- Refuses to rank schools "best to worst"
- Does not filter by demographic percentages
- Always includes poverty context (ENI) with performance metrics
- Presents multiple hypotheses for patterns, not single explanations
- Uses "schools facing challenges" not "failing schools"
- Acknowledges limitations in every response
Feedback & Questions
This is a portfolio project demonstrating AI-native data journalism. For questions about methodology or to report issues, please open an issue on the project repository.
View on GitHub