Algorithms The forbidding math of finding terrorists New York subway tests a gun detection technology
Cause-effect On the evidence raised by SAT distractors Kaiser reads a series of blog posts about SAT testing.
Cause-effect Other numbers from the colonoscopy trial Kaiser highlights some other metrics from the colonoscopy study
observational data With adjustments comes responsibility Kaiser explains how adjusted per-protocol analysis leads to different conclusions in the colonoscopy trial.
a/b testing Why no serious researchers conduct "per protocol" analyses Or, the importance of random treatment assignment