jstray / risk-ratios

Workbook to teach the concept of risk ratios for data journalism applications

risk-ratios

A notebook to explain and teach the concept of risk ratios for data journalism applications, created for a session at the NICAR 2022 data journalism conference:

What do vaccine effectiveness, pay-to-play meetings with politicians, employment discrimination, and TSA security screening have in common? They're all topics which rely crucially on a risk ratio, a simple formula which compares how often something happens to two different groups. While the math isn't particularly hard, many people don't know how to work with these ratios or spot the stories which depend on them, which results in a number of common reporting errors. We'll work together in a Python notebook to do the calculations behind all of the above stories, and learn what the risk ratio means in each context.

• Workbook with answers
• Workbook with blanks to figure out the answers here
• Slides that use pictures to teach about conditional probability and the base rate fallacy

Open source, CC-BY Jonathan Stray.