An Annotation for Transparent Inquiry (ATI) project analyzing perceptions of fairness in criminal justice decisions. The qualitative component focuses on a debate between ProPublica and Northpointe regarding the COMPAS risk assessment tool, while the quantitative component uses a dataset of over 7,000 individuals compiled by ProPublica. The data was gathered via Freedom of Information Act requests and analyzed using critical discourse analysis and SPSS.
Use Cases
- Analyzing media discourse and argumentation strategies based on the ProPublica and Northpointe documents.
- Triangulating themes and conclusions based on primary and secondary source materials.
- Studying perceptions of algorithmic fairness in risk assessment tools based on the described qualitative analysis.
- Conducting quantitative analysis on criminal justice data based on the ProPublica dataset of over 7,000 individuals.
Strengths
- Dataset includes over 7,000 individuals, providing a substantial quantitative sample.
- Primary source documents are publicly available, ensuring transparency and accessibility.
- Analysis employs critical discourse analysis and triangulation of multiple sources.
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download.
- Column-level documentation is absent; field semantics must be inferred after download.
- Access to some secondary source materials from Northpointe was limited due to trade secret claims.
Provenance
- Source
- ProPublica and Northpointe (COMPAS tool owner)
- Collection Method
- Data gathered via Freedom of Information Act requests; documents retrieved from corporate websites.
- Freshness
- Last updated 2026-05-04 07:13:16; freshness should be verified.