← All proof trees · SAMUEL POWEL SCHOOL

1. 1
   What you see3rd grade ELA proficiency (PSSA)

   73.5%

   the share of tested grade-3 students scoring proficient or advanced on the PSSA English language arts test. Based on 34 students.

   vs. peers

   96th percentile

   vs. citywide

   88th percentile

   unit

   percent

   direction

   higher is better

2. 2
   How it was computedvalue = published '% proficient or advanced' (accountability cohort by default)

   1. Unzip the district's PSSA/Keystone archive and read the accountability-cohort ('Acct') and all-tested ('Actual') school CSVs.
   2. Keep rows for the test, subject, grade, and subgroup this metric represents.
   3. Store the published '% proficient or advanced' as the value and the published number tested as the denominator; suppressed cells (asterisks/blanks) are stored as suppressed, not zero.
   4. Store the accountability-cohort cut as the default (population_cut = 'acct') and the all-tested cut alongside it ('actual').

   Citywide percentile, reconstructed right now

   Among 73 default-included schools with a value for this metric, year, and subgroup, this school has a higher value than 63. Midrank rule: (63 + 0/2) / (73 − 1) = 0.875.Stored: 0.875. ✓ matches

   How percentiles are computed · How peer groups are chosen

   Peer percentile, reconstructed right now

   Peer group: up to 40 of the demographically nearest schools among Elementary / catchment / district schools; 26 of 40 peers have a value this year. This school is above 25 of them. Midrank: (25 + 0/2) / (27 − 1) = 0.962; stored 0.962. ✓ matches

   The peers and their values (26)

   Peer school	Value
   A.S. JENKS ACADEMICS PLUS SCH	77.3%
   WILLIAM H. LOESCHE SCHOOL	52.1%
   ANNE FRANK SCHOOL	50.4%
   WATSON COMLY SCHOOL	48.6%
   RHAWNHURST SCHOOL	46.3%
   RICHMOND SCHOOL	46.1%
   FOX CHASE SCHOOL	43.6%
   J. HAMPTON MOORE SCHOOL	37.6%
   KENNEDY C. CROSSAN SCHOOL	32.1%
   JOHN M. PATTERSON SCHOOL	31.3%
   STEPHEN GIRARD SCHOOL	31.3%
   JOSEPH W. CATHARINE SCHOOL	30.8%
   ROWEN SCHOOL	29.2%
   HENRY W. LAWTON SCHOOL	26.8%
   JAMES R. LOWELL SCHOOL	26.7%
   EDWIN FORREST SCHOOL	25.4%
   FRANCES E. WILLARD SCHOOL	20.9%
   JOHN H. WEBSTER SCHOOL	19.5%
   JAMES J. SULLIVAN SCHOOL	18.5%
   WILLIAM CRAMP SCHOOL	18.5%
   FELTONVILLE INTERMEDIATE	11.9%
   JOHN B. KELLY SCHOOL	11.1%
   LEWIS ELKIN SCHOOL	8.9%
   LAURA H. CARNELL SCHOOL	8.3%
   THOMAS G. MORTON SCHOOL	8.0%
   ALEXANDER MCCLURE SCHOOL	4.7%

   Note: stored percentiles are computed by a batch script after loads; a difference from the live reconstruction means a newer load changed the comparison set and the batch has not been re-run.

   Evidence: metric tree: 3rd grade ELA proficiency (PSSA)

3. 3
   The stored cellThe stored cell

   table

   philly_school_year_metrics

   value

   73.53

   denominator

   34

   suppressed

   false

   citywide_percentile

   0.8750

   comparison_group_percentile

   0.9615

   source_load_id

   bba25088-d2c0-415a-b47e-c47d1dcb2edc

   Evidence: data model overview

4. 4
   The load that wrote itWritten on 2026-06-03 by sdp_pssa_keystone.ts

   source id

   sdp_pssa_keystone_2024-25

   file URL

   https://cdn.philasd.org/offices/performance/Open_Data/School_Performance/PSSA_Keystone/2024_2025_PSSA_Keystone_All_Data.zip

   SHA-256 of the file read

   d040e16c1b88f33dd481efbdb1a3924951ab7cc81869c8dee52458280574c13aCopy hash

   bytes

   705,262

   year covered

   2024-25

   code commit

   63ddc08c2a63170cecb6e78b991559cef9aa3926

   To confirm you have the same file: download it from the URL above, then run shasum -a 256 <downloaded file> Copy command — the output should equal the hash above. If the publisher has since replaced the file, the hashes will differ, which is itself useful to know.

   Evidence: loader script · first-fetch findings for this file · source registry

5. 5
   The publisher's filePSSA & Keystone results, 2024-25

   School District of Philadelphia (Performance Office). State test results by school, test, subject, grade, and subgroup: counts and shares at each performance level, in two population cuts — the accountability cohort (students enrolled for the full year) and all tested students.

   format

   ZIP of CSVs (accountability-cohort and all-tested cuts)

   join key

   id_eos (= ULCS Code)

   landing page

   https://opendataphilly.org/datasets/sdp-pssa-keystone-performance/

   • Two cuts of the same results; the accountability cohort is the canonical one served by default.
   • District-run and alternative schools only; charter results come from the state's Future Ready file.
6. 6
   How it was checkedChecks: 2 passed, 2 flagged

   • ✓ passedBase case: served cells vs an independent re-read of the ingested file
   • ✓ passedCompleteness: every source cell is served
   • ⚠ flaggedAccountability-cohort vs all-tested files are byte-identical after sorting in SDP's 2024-25 archive — a publisher defect, so this check cannot distinguish the two populations
   • ⚠ flaggedCross-publisher sanity check vs the state's Keystone results

   Checked at ±1: 0 mismatches, 0 cells served but absent from the file, 0 in the file but not served (ALL, acct, latest year).

   Evidence: Base case · Completeness · Accountability-cohort vs all-tested files are byte-identical after sorting in SDP's 2024-25 archive — a publisher defect, so this check cannot distinguish the two populations · Cross-publisher sanity check vs the state's Keystone results

7. 7
   Assumptions & caveatsAssumptions and caveats

   • The accountability cohort (students enrolled for the full year) is the intended canonical population. However, the two files SDP publishes for 2024-25 ('Acct' and 'Actual') are identical once sorted — 28,525 rows with no differing value or denominator — so the served 'acct' and 'actual' cuts carry the same numbers. We treat this as a publisher defect (one file published under two names) and cannot say which population the numbers describe.

   • District-run and alternative schools only from this file; charter schools' Keystone results come from the state's Future Ready file.

   Algorithms involved: Citywide and peer-group percentiles, Peer groups: 40 nearest neighbors on demographics