Auditing mortgage-denial disparities

Published by Emmanuel Martinez and Lauren Kirchner / The Markup

This credits the original publisher. Better Loop membership or a shared assessment is not implied.

The public work

The Markup analyzes 2019 mortgage-application records with logistic regression and publishes adjusted disparities. Its methodology documents selection rules, model variables, missing credit-score information and limitations of the available public data.

What to notice

Document selection rules and unobserved variables before drawing conclusions; an adjusted association does not reveal every mechanism behind a decision.

Keep the context

Observational analysis of a selected historical loan population. Unavailable credit scores and other omitted information limit explanations. Adjusted associations do not establish the cause of an individual decision.

AI use: Not reported in the source.

The inspected source does not report AI-assistant use by its authors.

A useful public example is not an assessment of a reader, a publisher or a Better Loop member.

Authored practice suggestion

Try the idea.
Check your own work.

Use material you are allowed to work with. This suggestion is preparation; it does not record a completed task or an improvement.

A prompt to adapt privately

Using a local copy of the published methodology, construct an evidence map covering population filters, outcomes, predictors, missing variables and robustness questions. Trace one reported aggregate result to its defined comparison. Write a conclusion that separates an association from a causal explanation.

A check to adapt

The comparison population and outcome are explicit; missing credit scores are acknowledged; odds ratios are not mislabeled as raw denial-rate differences.

Open a related practice brief

Private practice needs no account. Your configured model provider may process the evidence you select.

See the full skill-to-sharing flow

Another way into the work.

Mapping grocery proximity in Chicago

The guide projects Chicago community areas and grocery locations into a metric coordinate system, buffers stores by one kilometre, and displays intersection and difference maps. It defines coverage geometrically.

Published byGeoPandas developers

Curated public exampleNotebook

Everyday work

Compare audio and visuals before summarizing video

OpenAI’s notebook processes a real DevDay recap into frames and audio, then prints summaries and answers from visual input, audio-derived text, and both together. The differing outputs make missing context visible.

Published byOpenAI Cookbook / Juston Forte

Curated public exampleReport

Research & strategy

Accounting for food-system emissions

Hannah Ritchie’s analysis visualizes food emissions using Poore and Nemecek’s published meta-analysis. It separates livestock, crops, land use and supply-chain contributions, explaining category boundaries and the distinction between food-system and global totals.

Published byHannah Ritchie / Our World in Data