NJ AP Research
Independent research · public New Jersey school data

Understanding Advanced Placement opportunity in New Jersey Public High Schools, one pipeline stage at a time

A linked suite of empirical, methodological, and administrative studies that measure how formal AP course access, examination participation, conditional success, and overall successful attainment relate to one another — statewide, and school by school — and turn that evidence into bounded school-level reports that translate rather than refit the underlying research.

The A → R × C pipeline

Four indicators, one identity, kept deliberately distinct so that no single statistic has to describe the whole AP program.

AP Access (A) Reported formal AP course enrollment among the relevant Grade 11–12 population Examination Reach (R) AP examination participation relative to the relevant student population Conditional Attainment (C) Successful AP attainment conditional on reaching the examination stage × Successful Attainment (S) Successful AP attainment relative to the full relevant population — by identity, S = R × C
AP Access and Examination Reach describe opportunity and participation; Conditional Attainment describes performance once a student has already tested. Successful Attainment is not a fifth, independently measured quantity — it is the product of Reach and Conditional Attainment, which is why two schools can share a Successful Attainment rate for very different reasons. The analytic cohort is 276 AP-aligned comprehensive public high schools with no observed IB offering in the governed support window.

Program architecture

Seven linked projects, moving from statewide measurement to institutional use. The program is cumulative, not linear — several studies share data and constructs while asking different scientific questions.

1. Statewide AP-Aligned Study 2. Inside Successful Attainment 3. Exam Reach AP-Aligned Study 4. AP Access Attainability Study 7. Stability & Directionality planned 5. Administrative Report Engine operational 6. Coefficient Arithmetic planned
Solid arrows carry a direct evidentiary or operational linkage; dashed arrows are a validation linkage; the dotted line is a conceptual linkage. Stability & Directionality validates the single global Reach deviation inherited from Inside Successful Attainment — it is not a validation branch of the Report Engine's Reach diagnostic, and Coefficient Arithmetic does not re-fit any downstream model.
direct / operational validation conceptual

Seven studies, four roles

Status reflects where each study actually stands today. This page updates as findings, papers, and reproducibility packages become ready — nothing here is a placeholder for a claim we haven't made yet.

Featured findings

From the program's completed empirical foundation, the Statewide AP-Aligned Study.

Statewide AP-Aligned Study · 276 schools · 1,375 school-year observations · 2020–21 through 2024–25

Broader AP access moves with more successful attainment — even as conditional pass rates soften

As formal AP Access grows within a school, do Examination Reach, Conditional Attainment, and Successful Attainment move together, or can broadening opportunity coexist with lower pass rates among the students who test?

276 AP-aligned comprehensive public high schools, no IB offering
1,375 school-year observations, 2020–21 through 2024–25
0.34 → 0.32 pp Successful Attainment gain per 1-pp Access increase: baseline → composition & absenteeism–adjusted
  • A 1-percentage-point within-school increase in reported formal AP enrollment is associated with a 0.34-percentage-point increase in Successful Attainment in the baseline school-and-year fixed-effects model, and 0.32 points once student-composition and chronic-absenteeism controls are added.
  • That enrollment increase is strongly associated with more Examination Reach but negatively associated with Conditional Attainment, the pass rate among students who actually test; annual-change models show the same directional pattern. Broader opportunity can coexist with a lower conditional pass rate while still raising the share of the full cohort reaching a successful outcome.
  • Most of the gap between examination reach and reported enrollment is a stable, between-school pattern rather than something that emerges within a school over time, and the enrollment–attainment coefficient gets slightly larger, not smaller, when the most discordant schools are excluded or trimmed.
  • The relationship is associational, not causal: it describes a statewide empirical regularity, not a guaranteed effect of expanding access at any one school.
Downloads

School Reports

From the Administrative Report Engine — a downstream layer that translates rather than refits the underlying research. It has no paper or findings of its own; it assembles governed evidence from the studies above into a bounded report.

Reach and Access evidence are evaluated independently under pre-specified, governed rules. A school that clears both gets the integrated AP Reach & Access Opportunity report; a school that clears only the Access rule gets the AP Access Opportunity report; a school that clears neither is routed to review and receives no report.

25 Reach-qualified and Access-qualified — eligible for the integrated AP Reach & Access Opportunity report
122 Access-qualified, Reach not qualified — eligible for the AP Access Opportunity report
8 Reach-qualified, Access not qualified — reach opportunities, currently routed to review with no dedicated report branch

Counts generated by Report Engine v1, run August 31, 2026 against 2024–25 reference-year data across the governed 276-school AP-aligned cohort (33 meet the Reach Rule and 147 meet the Access Rule, 147 = 25 + 122). These are a snapshot, not a live count — they will move as later school-years and engine versions are run.

How report qualification works

Reach Rule

Consumes the pinned AP Examination Reach study outputs unchanged. A school passes when all of the following hold:

  • the latest-year universal-prediction (U) Reach discrepancy is positive;
  • its current U absolute-error percentile is at or above the 75th;
  • at least 3 of its annual U references are interpretable;
  • at least 75% of those interpretable years show a positive discrepancy;
  • the current universal (U), locale-contextual (L), and local structural (G) discrepancies are all positive; and
  • the current universal prediction falls inside the governed model domain.

Access Rule

Consumes the AP Access Attainability Study's governing Prior-Preparation geometry through a downstream adapter that consumes, but never recalculates, the upstream model. A school passes when the metric support distance to its Kth persistent supporter, δ(h,K), is at or below dmax — under the engine's default parameters:

  • h = 5.0 percentage points — the minimum contemplated persistent Access increase;
  • K = 5 — the required number of distinct persistent supporting schools; and
  • dmax = 0.50 — the maximum standardized comparability distance to that 5th supporter.

Changing h, K, or dmax changes only downstream routing — never the upstream Reach or Access science it draws on.

Two sample reports illustrate the content and format:

About the program

Independent research using publicly available data. NJ AP Research is not affiliated with, sponsored by, or endorsed by the New Jersey Department of Education or the College Board. The program is diagnostic, not an accountability rating: it does not rank schools, does not treat structural deviations as proof of school quality, and does not reverse-engineer a target from a desired outcome.
Primary data source
New Jersey Department of Education, School Performance Reports (downloadable school-level files).
Scope
276 AP-aligned New Jersey comprehensive public high schools; formal AP Access, Examination Reach, Conditional Attainment, and Successful Attainment.
Contact
Dr. Yang, Yi — theyangs20@gmail.com — questions, corrections, and methodological feedback welcome.