The Anderson School scores at the ceiling — and proves nothing

Anderson is 40 to 50 percentage points above the typical New York City elementary school across every grade and subject, in both math and reading. The pattern is flat across the grades, which rules out grade-band specialization as the explanation. Almost every Anderson student is proficient on almost every test.

522 students. 9% economically disadvantaged (city avg ~70%). 43% white, 32% Asian. Anderson admits via G&T testing in kindergarten — kids enter scoring at the 90+ percentile of the NYC Gifted & Talented assessment. The school selects for kids who would score above proficient at any school in NYC.

Anderson's 98% proficient rate is essentially what you'd predict from its admission profile. The students would likely score 95%+ proficient at a number of New York City schools — so for the proficiency-rate question, Anderson doesn't differentiate itself. That doesn't mean Anderson isn't adding anything. What it means is that proficiency on a third- through seventh-grade state test is the wrong metric. The students cleared that bar in kindergarten. To see what Anderson is contributing on top of that, you'd want measures further up the curve: depth of math curriculum, share of students reaching advanced levels, downstream high-school placement, and qualitative measures of teaching practice.

For a parent: Anderson offers small classes, high resources, and a peer environment of students all working at or above grade level. For a researcher: the proficiency rate is a low-information metric here, but the school is still a case study — in selection effects, in how a high-input cohort interacts with instruction, in the long-tenured leadership that's kept the program stable. New York City has very few K-8 Gifted & Talented programs, Anderson is the most-coveted, and admission depends on the kindergarten Gifted & Talented test (a test that families with the resources to prepare for it are over-represented on). Both of those facts shape who attends.