Method and limits

Words this data uses

New York City's school data has a vocabulary of its own. None of it is hard, but all of it is assumed.

DBN

District, borough, and school number, such as 02M475. Every City school has one and it never changes, which is why everything on this site is joined by it rather than by name. The first two digits are the district, the letter is the borough, the last three are the school.

Comparison group

A set of schools the City treats as similar to this one when judging a measure. The City chooses the membership and publishes the group's average, but not the list of schools in it, so neither this site nor you can see exactly who is in a given group.

Score out of 5

The City's own rating of one measure against that comparison group. Higher is stronger. It is not a rating of the school.

Withheld

The City calculated a value and did not publish it, because so few students are in the group that publishing it could identify them. It is a protection, not a bad result, and it is not a zero.

Economic Need Index

The City's estimate of the share of a school's students facing economic hardship, based on public assistance, housing and similar records rather than on a survey. Higher means greater need.

Report type

Which quality report a school files: elementary and middle, high school, transfer school, early childhood, District 75, or Young Adult Borough Center. It decides which measures a school publishes at all.

Screened, zoned, open

Admissions methods, set per program rather than per school. Open means no academic requirement. Zoned gives priority to students living in the school's zone. Screened uses grades or other criteria.

What this site is

New York City publishes a great deal about its schools, in several places that do not join up. A school's test results sit in one dataset, its enrollment in a workbook, its address and admissions in a third file released once a year for one admissions season. This site puts them together on one page per school and states, for every value, what it measures, which year it describes, and where it came from.

It is a reference, not a recommender. There is no overall score, no ranking and no weighting. Those require deciding what matters, and that is not a decision a website should make for a family.

Which schools are included

A school appears here if New York City Public Schools publishes data about it: a quality report, an enrollment snapshot, or both. That is every district school, including District 75 special education schools, District 79 alternative programs, transfer schools and Young Adult Borough Centers.

The admissions directories are used to describe a school, never to admit one to the list. They also contain early childhood centers whose codes are not school numbers, and they cover one admissions season rather than every school that exists.

A school is marked open when it appears in the newest enrollment snapshot or in a current directory. Everything else is marked closed or former and keeps its full history. Neither label comes from an official status field, because none of the sources publishes one. It is this site's reading of the evidence, and it can be wrong for a school that has just opened.

Joining the sources

Everything joins on the DBN, the district, borough and school number that identifies a school across every city system. Nothing joins on a school name. Names are spelled differently in every file, and matching on them quietly attaches one school's results to another.

Borough and district are read from the DBN itself rather than from a lookup file, so they cannot go stale.

Missing, withheld, and not applicable

These are three different facts and the site keeps them apart.

• Not reported means the source published no value for that school and year.
• Withheld means the source deliberately suppressed the value because too few students were in the group. The measure exists. The value is protected. It is not a zero and not a low score.
• Does not apply means the measure is not published for that type of school. A graduation rate is not missing from an elementary school. It was never expected. Those measures are left out of a profile rather than listed as gaps.

No value on this site is ever filled in, estimated or defaulted to zero. A figure that could not be read from a source is absent.

Why an absence is never shown as zero

A zero and a blank look similar and mean opposite things. If a school has no Native American students, the enrollment snapshot says 0.0% and this site shows 0.0%, because that is a measurement. If the City withheld a test result for the same group, writing 0% would state that none of those students met the standard, which nobody has said and which is probably false. So a withheld figure says withheld.

Every group is listed, every time

Under a measure, the same list of student groups appears for every school: each race and ethnicity, each gender, English language learners, students with disabilities, and each grade the measure covers. A group with no figure is listed with the reason rather than left out.

Leaving it out was the earlier behavior and it was worse than it looked. A race breakdown would show two groups and silently omit the rest, and there was no way to tell from the page whether the school had no such students, the City had withheld the figure, or the site had a hole in it. In one real case the answer was that four of the five missing groups had been withheld, which the City had said plainly and the page had thrown away.

When two numbers should not be compared

Different years. Schools do not all report in the same year. A comparison row states the period, and says explicitly when the two schools' most recent values come from different years.

Across 2019-20 and 2020-21. State testing was canceled in spring 2020 and participation was unusually low the following year. Attendance in those years counts remote days. A trend line through that period is measuring the pandemic, not the school.

Renamed measures. A few measures were redefined between the 2017-18 and 2018-19 school years and given new identifiers. They are published here as separate series with a note, rather than stitched into one line that would imply continuity.

Different school types. A school serving grades 6 to 12 files two quality reports, one for its middle grades and one for its high school grades, and publishes some measures in both for different students. Those appear as two labeled rows, not one.

Different group sizes. A percentage over 12 students and a percentage over 900 are not equally stable. Where the source publishes the group size, it is shown next to the value.

Sorting a comparison. The comparison table can be sorted by any column. That reorders a list by one published measure in one year, which is a legitimate thing to want. It is not a ranking of the schools, and sorting by a different column will often produce a different order.

Comparison groups, scores, and the colored bands

A raw percentage is hard to read on its own. Seventy-one percent at Level 3 or 4 in English is good or unremarkable depending on the school, and nothing on the page can tell you which without a reference point.

New York City provides that reference point itself. For many measures it publishes two extra figures alongside the value: the average for the school's comparison group, and a score from 1 to 5 reflecting how the school did against that group. Both are the City's own calculations, made against a set of schools the City considers similar to this one.

This site uses them in three ways, and adds nothing of its own:

• The comparison group average is written out as a sentence rather than a second bare number, so you see the distance and its direction: five points above a 89.2 percent average, not just "89.2".
• Where the City publishes a score, the value carries a colored band. The score itself is printed inside the band, out of 5.
• The band always carries a written label as well as a color, because a color alone is not readable by everyone and is not a reason on its own.

The four bands are this site's grouping of the City's number: 4 and above, 3 to 4, 2 to 3, and below 2. That grouping is the only judgment involved, it is stated here, and the underlying score is always visible so you can disagree with the grouping and still read the number.

Where the City publishes no score, nothing is colored. That is common: roughly one published value in six carries a score, and demographic figures never do, because there is no better or worse version of a school's enrollment.

A band is not a rating of the school. It describes one measure, in one year, against one comparison group chosen by the City.

One year to a measure

Every figure in a measure comes from the same school year: the year its all-students figure was last reported. A student group with nothing that year reads as not reported rather than falling back to an older number.

This matters more than it sounds. Left to themselves, groups drift to their own most recent figures, and a card ends up showing every group at 100 percent and English language learners at 85.7, with the years eight apart in small grey type beside them. That is not a hard comparison to misread. It is an impossible one, and it looks like a fact about the school rather than an artifact of the page.

Earlier figures are not lost. Every measure's panel carries the full year-by-year series.

How measures are ordered

Within a section, a measure leads with its all-students figure, and any breakdowns by student group sit under it. Those groups are shown in a stated order: race and ethnicity, then gender, then student groups such as English language learners, then service settings, then prior achievement. Within each of those, the order is alphabetical.

Alphabetical is deliberate. Any other order inside a list of racial or ethnic groups is a judgment about which group to read first, and that is not a judgment this site has any basis to make.

Group and measure names are the Department of Education's own wording, copied rather than reworded. Where the source writes "Multi-Racial" or "Economic Need Index", so does this site, because those are the terms you will meet again in every other City document about the same school.

Units

Most percentages arrive as proportions between 0 and 1 and are shown as percentages. Proficiency values are a scale of roughly 1.0 to 4.5, not a percentage. Where a source does not state a unit, this site infers it from the range of published values and says so in the data dictionary and on the measure itself. An inferred unit is a judgment, not a fact from the source.

Addresses and coordinates

Addresses come from the current admissions directories. High school coordinates are published alongside the address by the source. For every other school the coordinate is matched from the address using the New York City Department of City Planning geocoder, which makes it a derived value. Each school records which of the two it has. An address that does not match produces no coordinate rather than an approximate one.

About four hundred open schools have no published address, because they do not appear in an admissions directory. Their profiles show no address section rather than a placeholder.

This version of the site does not embed a map or load a mapping library. An address links out to a map instead.

Programs and admissions

Admissions happen at the program level, not the school level. One school can run an open program and a screened one at the same time. Seats, applicants and priority order are shown per program and describe the Fall 2025 admissions season only. They are a record of one season, not a prediction about the next.

Freshness

Every reporting period on this site is read from inside the data. It is never the date the build ran, and never the date a source page says it was updated. Several New York City open data pages report a recent update when only the description changed; at least two datasets that describe themselves as annually updated last received rows in 2011 and 2013, and they are not used here.

When a source passes its expected refresh window the site shows a warning on every page. Silent staleness would be worse than a missing figure.

What is not here

• No overall score, rank, or recommendation.
• No commute or travel time.
• No address entry, zoning lookup or interactive map.
• No survey results yet. Those live in workbooks not used in this version.
• No SHSAT or specialized high school admissions figures. The high school directory carries them in a separate set of columns whose grain is not documented clearly enough to publish safely.
• No charter schools beyond what the city's own reporting covers.

Language

This site is in English only. The City publishes school information in nine languages and its MySchools service is translated; if English is not your first language, that is the better starting point for applying. A translated version of this site is a larger piece of work than one release, and running the pages through machine translation would produce confident-sounding errors about a child's education, which is worse than saying plainly that it has not been done.

Errors

If a figure looks wrong, it is worth checking against the source linked on the measure. If the error is here rather than there, report it.