Register once, download everything
We don't subscribe you to any list. No spam, no newsletters, no follow-ups. We collect your email and zip code only to understand where these materials are being used and to measure the project's reach.
Downloads unlocked. You're registered — download any materials below.
1
The Algorithm Has an Opinion
What algorithmic bias is, how it enters healthcare AI, and why a system can produce biased results with no biased person anywhere in it. Students trace one modern disparity back to its documented historical origin.
Covers difficult medical history (Tuskegee, Sims, forced sterilization) with a built-in content-warning protocol.
2
Skin Deep — When AI Can't See You
Medical devices and AI systems that perform measurably worse on darker skin — pulse oximetry, dermatology AI, newborn jaundice screening — and the training-data pipeline that produced them. Students run their first training-set audit.
3
The Data Gap — Who Gets Counted
If you were never in the study, the medicine was not designed for you — and nobody had to intend that. Students investigate which populations are overrepresented in medical research, which are missing, and what happens in the gap.
4
Race in the Formula
For decades, medical algorithms used race as a variable — affecting kidney function estimates, lung capacity readings, and heart risk scores. Students read the formulas themselves and argue both sides of removing race from clinical algorithms.
5
Your Health, Your Data, Your Rights
The data that trains medical AI comes from people — most of whom never knowingly agreed. What HIPAA actually covers, why "de-identified" is a weaker promise than it sounds, and who currently holds your health data. Students audit a real app's privacy policy.
6
From Diagnosis to Action
The capstone: teams run a structured bias audit on a real, cited entry in the diagnosethebias.org database — including pulling a citation and verifying it says what the entry claims — then present a one-page bias brief with one concrete remedy.
Capstone requires student devices with internet access for the live database audit. (Sessions 1-5 run without student devices.)
Creative Commons BY-NC-ND 4.0
All Diagnose the Bias curriculum materials are licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
You are free to share these materials, but you must give appropriate credit, you cannot use them for commercial purposes, and you cannot distribute modified versions.