Vaccine researchers discard the start-at-zero rule
Kaiser revisits the practice of manipulating the vertical axis of a column chart.
Statistics
One-dose Pfizer is not happening and here's why
Kaiser explores what the Pfizer vaccine trial data say about a one-dose treatment.
Tautology in data arguments
Kaiser discusses statistical tautologies.
Why tests are not perfect
Kaiser explains why testing can't be perfect. Typically, the test itself can't be perfect; but also errors arise from outside the labs.
In times of paranoia, seek out data and stick to facts
Kaiser Fung (Principal Analytics Prep, Numbersense) urges people to stick to the facts as they assess the coronavirus public-health situation.
Good models + Bad data = Bad analysis
Example showing how to diagnose bad data in data science models