Source document

docs/design/reference-scaffold/config/comparability.yaml

Served verbatim from the project repository. Internal working document conventions apply: documents may reference file paths, branch names, and findings-ledger anchors from the repo.

# Comparability regimes. Trends are computed WITHIN a regime, never across a break.
# Because View 2 trends run on within-year standardized residuals (z), the trend is
# already scale-break-immune *and* demographic-adjusted; regimes additionally prevent
# fitting a slope across a structural discontinuity (e.g. the 2023 NextGen standards reset).
regimes:
  ela_math_regimes:
    description: >
      NY grades 3-8 ELA/Math. 2023 NextGen standards reset makes proficiency
      non-comparable pre/post 2023 (NYSED + NYCDOE state this directly). 2020 cancelled;
      2021 had ~21% participation (unusable for school-level rates).
    windows:
      - {id: A_commoncore, years: [2013, 2014, 2015, 2016, 2017, 2018, 2019]}
      - {id: B_nextgen,    years: [2023, 2024, 2025]}
    excluded_years: [2020, 2021]
    ambiguous_years: [2022]                 # old standards, recovery year
    cross_regime_reporting: standardized_only

  chronic_absenteeism_regimes:
    description: COVID structural break; post-pandemic levels are elevated citywide.
    windows:
      - {id: pre_covid,  years: [2016, 2017, 2018, 2019]}
      - {id: post_covid, years: [2022, 2023, 2024, 2025]}
    excluded_years: [2020, 2021]
    ambiguous_years: []
    cross_regime_reporting: standardized_only

  survey_regimes:
    description: >
      Placeholder. Set instrument-stable window(s) only AFTER verifying NYC School
      Survey item wording/scale continuity across years for each survey outcome.
    windows:
      - {id: TBD, years: []}
    excluded_years: []
    ambiguous_years: []
    cross_regime_reporting: standardized_only