How to read NYC school data honestly

Numbers feel objective, so it is tempting to rank schools by whichever stat is biggest. But every school metric is an answer to a specific question, and the question is often narrower than the headline suggests. This guide is about reading the numbers honestly: what each one actually measures, where it quietly misleads, and why the highest-scoring school on a list may still be the wrong school for your child.

How it works in 60 seconds

• Graduation rate and college readiness are different bars. A diploma is not the same as being ready for college-level work, readiness is the harder, more telling number.
• A school-wide applicant-to-offer ratio is not your personal odds. Your real chance depends on your priority group, the admissions method, and your random number.
• High scores can reflect who got in, not what the school added, that is selection bias, and it is strongest at screened admission and specialized schools.
• Small programs are noisy, charter cohorts can shrink, and private schools rarely publish comparable data, each is a reason to read a number with care.
• Quality is not fit. A genuinely strong school can still be wrong for your child on commute, environment, or program match.

A simple habit covers most of this: for every number, ask what exactly did this count, for whom, and in what year?

Ratios are not your odds

The most over-read number in NYC admissions is the applicant-to-offer ratio. It is genuinely useful as a demand signal, how popular a program is, but it is not your personal probability of getting in.

A school-wide ratio averages over every kind of applicant at once. Your actual chance is shaped by three things the ratio hides:

1. Your priority group. Higher tiers, for example zoned or sibling priority, are considered before lower ones. A program that looks long-shot citywide can be near-certain for an in-group applicant, and the reverse.
2. The admissions method. A lottery, a screened admission program, and an audition each turn the "same" ratio into very different odds for the same child.
3. Your random number. Within a priority tier, ties are broken by the single random number you receive each cycle, so two identical applicants can land on opposite sides of the cutoff.

There is one more split parents miss: general-education and students-with-disabilities (SWD) seats can have different ratios. Programs often report seats and demand separately for each track. If your child needs the SWD track, the GenEd ratio is the wrong number, and vice versa. Always read the ratio for the seat type that applies to your child.

The honest takeaway: use ratios to gauge how contested a school is, then get your personal odds from a tool or counselor that accounts for your priority, method, and seat type, not from the headline number.

Survey vs. outcomes, selection bias, and small cohorts

Different datasets answer different questions, and mixing them up is the most common reading error.

Survey data vs. outcomes data. Family, student, and teacher surveys measure experience and perception, whether people feel the school is safe, supportive, well-run. Outcomes data, test scores, graduation, readiness, measure results. A school can score high on one and low on the other. A warm, well-loved community is not automatically a high-results school, and a high-results school is not automatically a happy one. Read surveys for climate and outcomes for results, and never let one stand in for the other.

Selection bias in test scores. This is the big one. Screened and specialized schools admit students who were already high-performing. So strong average scores there partly measure who walked in the door, not how much the school added while they were there. The question that cuts through it is: did students grow, or did the school simply start with students who would have scored well anywhere? Where a "value-added" or progress measure exists, it is far more revealing than a raw average. A high raw score at a screened school and the same score at an open-admission school do not mean the same thing.

Small-cohort noise. A tiny program, a handful of graduates, a small testing cohort, produces numbers that swing wildly year to year for reasons that have nothing to do with teaching. One unusual class can move a small school's "percentage" dramatically. With small cohorts, prefer multi-year patterns over a single year's figure, and treat a dramatic one-year jump or drop as possible noise until a trend confirms it.

Charter attrition and private-school data gaps

Two structural blind spots deserve their own warning, because they distort comparisons in opposite directions.

Charter attrition and backfill. A charter cohort can shrink as it moves up the grades, and a school is not always required to refill, "backfill", the empty seats. If students who leave are not replaced, the cohort that takes the upper-grade tests or graduates can look stronger than the group that started, without the school having changed anyone's trajectory. This is not an accusation, policies vary by school, but it is a reason to ask: is this the same cohort that started, and were leavers backfilled? Compare entering-grade and exiting-grade enrollment before reading upper-grade outcomes at face value.

Private-school data gaps. Private and independent schools rarely publish the standardized metrics the public sectors report, there is often no comparable graduation, readiness, or test-score line at all. That is a data-availability issue, not a quality signal. A blank where a number would be does not mean the school is weaker; it means the number was never required to be published. Judge private schools on what they do disclose plus your own visit, and resist reading the missing data as either a strength or a weakness.

High-performing is not the same as right-for-your-child

Even after you have read every number correctly, the highest-scoring school on your list can be the wrong choice, because quality and fit are different things.

A school's outcomes describe its students in aggregate. Fit describes the match between this school and your child:

• Commute. A long, hard daily trip taxes a child every day and can undo the advantage of a stronger school.
• Environment. Size, structure, pace, and culture suit different children differently, a high-pressure setting that lifts one child can sink another.
• Program match. A school can be excellent overall and still lack the language program, support services, or focus your child needs.

None of that shows up in a graduation rate. A school can be genuinely high-performing and a poor fit for a specific child, and a more modest school on paper can be the better place for that child to thrive. That is why we score fit to your child rather than rank schools by a single "best" number: the goal is the right school, not the highest-scoring one.

Does this apply to your child?

Official source

Common questions

If a school admits 1 in 5 applicants, are my odds one in five? Not necessarily. A school-wide ratio mixes everyone together. Your real chance depends on your priority group, the admissions method, and your random number, and the seat ratios for general-education and students-with-disabilities seats can differ, so read the one that applies to your child.

A school has top test scores, does that mean it teaches well? Not by itself. Screened and specialized schools admit students who already test high, so strong scores partly reflect who got in, not how much the school added. Look for whether students grew, not just where they started.

Why do we show less data on private schools? Private schools rarely publish the standardized metrics public sectors report, so the gap is about data availability, not school quality. Absence of a number is not evidence of weakness.