← All proof trees · MURRELL DOBBINS HIGH SCHOOL

1. 1
   What you seeOSS — % students with any suspension

   20.1%

   the share of a school's students with at least one out-of-school suspension in the year. Based on 700 students. For this measure, lower is better — percentiles below are flipped so that a high percentile is always the good outcome.

   vs. peers

   18th percentile

   vs. citywide

   8th percentile

   unit

   percent

   direction

   lower is better (percentiles are flipped so a high percentile is always good)

2. 2
   How it was computedvalue = 100 − published '% with Zero OS Suspensions'

   1. Read the district's suspensions CSV; keep directory schools and mapped subgroups.
   2. Derive the share with any suspension as 100 minus the published zero-suspension share.

   Citywide percentile, reconstructed right now

   Among 215 default-included schools with a value for this metric, year, and subgroup, this school has a better (lower) value than 18. Midrank rule: (18 + 0/2) / (215 − 1) = 0.084.Stored: 0.084. ✓ 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 High / lottery / district schools; 11 of 11 peers have a value this year. This school beats 2 of them. Midrank: (2 + 0/2) / (12 − 1) = 0.182; stored 0.182. ✓ matches

   The peers and their values (11)

   Peer school	Value
   CONSTITUTION HIGH SCHOOL	6.4%
   SWENSON ARTS/TECH HIGH SCHOOL	8.3%
   THE U SCHOOL	8.6%
   PAUL ROBESON HIGH SCHOOL	9.9%
   HIGH SCHOOL OF THE FUTURE	11.5%
   BUILDING 21	13.1%
   THE WORKSHOP SCHOOL	14.1%
   THE LINC	15.1%
   JULES MASTBAUM HIGH SCHOOL	15.2%
   RANDOLPH TECHNICAL HIGH SCHOOL	20.4%
   PHILADELPHIA MILITARY ACADEMY	24.3%

   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: OSS — % students with any suspension

3. 3
   The stored cellThe stored cell

   table

   philly_school_year_metrics

   value

   20.14

   denominator

   700

   suppressed

   false

   citywide_percentile

   0.0841

   comparison_group_percentile

   0.1818

   source_load_id

   9ed07178-853d-4052-aa83-1bfe11ce1d24

   Evidence: data model overview

4. 4
   The load that wrote itWritten on 2026-06-04 by sdp_attendance_and_suspensions.ts

   source id

   sdp_suspensions_school

   file URL

   https://cdn.philasd.org/offices/performance/Open_Data/School_Performance/Suspensions/SDP_OSS_School_S_20260213.csv

   SHA-256 of the file read

   6eaa8e3907d80e2572464dbcef793fafb312567cdfa120f3a4820f08fca517c4Copy hash

   bytes

   2,057,896

   year covered

   —

   code commit

   f75e33137e8166eec3f8b5a4b39463d306ef710c

   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 fileOut-of-school suspensions, by school and year

   School District of Philadelphia (Performance Office). For each school and year, the share of students with 0, 1, 2, and 3 out-of-school suspensions, by subgroup.

   format

   CSV

   join key

   ULCS Code

   landing page

   https://opendataphilly.org/datasets/sdp-school-performance/

   • No published '4 or more' column — derived as a residual (see the suspensions algorithm).
6. 6
   How it was checkedChecks: 1 passed

   • ✓ passedComputed-value check: 8,034 of 8,034 derived cells match an independent recomputation

   Evidence: Computed-value check

7. 7
   Assumptions & caveatsAssumptions and caveats

   • The published percentages for 0/1/2/3 suspensions are of the same denominator ('Total Students (Yearly)').

   • District-run schools only.

   Algorithms involved: Suspension shares derived from the published 0/1/2/3 columns, Citywide and peer-group percentiles, Peer groups: 40 nearest neighbors on demographics