Count NYC Dog-Name Records and Explain Your Denominator: A Reproducible Data-Literacy Lab

Abstract

A reproducible teaching lab about missing-name policies, uncertain strings, tied ranks and denominator choice in NYC DOHMH dog-license records. The archive contains the lesson and facilitator answers, a saved two-field public-data projection, Python standard-library reproduction and verification scripts, calculated CSV/JSON tables, and a fictional renewal scenario. The 10 October 2026 retrieval preserves 819,323 rows with only animalname and extract_year. Its 2026 extraction slice has 96,459 rows, a baseline denominator of 93,902 and a sensitivity denominator of 93,241. Rows are active license periods, not confirmed unique dogs, annual puppy registrations or a citywide population census. The source update date remains 15 July 2026. The scripts run offline; no classroom or human-learner test is claimed. Original resource and teaching context: iDogNames original lesson and resource . Underlying source: NYC DOHMH via NYC Open Data, nu7n-tubp. Original teaching material and generated tables are CC BY 4.0; Python scripts are MIT. City source records, metadata and attached City documents retain their own source conditions. License scopes, attribution and the City disclaimer are preserved in LICENSE.md and reuse.txt. This is an independent resource with no City endorsement. AI assistance was used to create and archive the lesson and code.