How the adjusted IAR view works

Raw test scores mostly tell you who a school admits. The adjusted view asks a different question: how does a school perform relative to what its demographics predict? A school beating expectation while serving 75% low-income students may be a stronger teaching unit than a higher-scoring school in a low-poverty neighborhood - and a family deserves to see that signal, honestly bounded.

The method, in plain words

Across all 468 CPS elementary and middle schools with published data (Illinois Report Card, spring 2025), average IAR proficiency falls by about 0.6 percentage points for every additional percentage point of low-income share (R² ≈ 0.61 - this one factor explains roughly 61% of the variation between schools). We fit that line excluding selective-enrollment and gifted schools, whose admissions filters would distort it. A school’s adjusted score is its actual score minus what the line predicts for its low-income share, re-centered on the district average - so a value near the district average means “performing exactly as expected,” and higher means beating expectation.

How to read it

We show adjusted results as rank movement and cohorts, not decimal points: “If all schools served the same share of low-income students, this school would rise two places among your candidates and sit in the middle 50% of CPS schools with published results.” Small point differences are noise; movement and company are the honest signal.

Why % IEP is not in the model

We tested a second adjustment for the share of students with IEPs and publish the result so you don’t have to wonder: it is a weak explainer (R² ≈ 0.08 on its own; adding it to the low-income model improves fit by only ~2%). Just as important, several top-rated schools - Lenart, Keller, and McDade among them - suppress their % IEP figure, so an IEP term would silently drop exactly the schools families ask about most. Where a number is suppressed we say “unknown” - we never estimate it.

Limits you should know

• School-level only. This says nothing about any individual child; a school-level pattern must never be read as a prediction for your kid.
• Not causal proof. Unmeasured differences (programs, stability, funding, community) ride along with the low-income share; “beats expectation” is evidence about the school’s contribution, not a verdict.
• Selective schools are flagged, never adjusted. Their gap above expectation is admissions selection, not a school effect.
• Within one year only. Illinois changed IAR cut scores in 2025; we refit every year and never compare adjusted values across years.
• Raw scores stay visible. Adjusted never replaces raw - you choose which one feeds your own weighting.

Coefficients are refit annually on the newest Illinois Report Card public data set; the current fit is stable across 2024 and 2025 (slopes within ~4%). This view is SchoolNav’s own analysis of official public data - it is not a CPS or ISBE rating. Back to Compare & Rank