Of General Statistical Interest

Statistical Evidence The solution to the replicability crisis is not to play a math game (i.e. to lower a p-value threshold) but to report effects that are clinicaly relevant. See …

  1. Slides on false discovery and confirmation rates.

  2. ROPE in sequential testing (See slides at ~5:45m on false discovery rates. Also, here.)

Statistical Thinking Insights on inference, clinical trials, and, well, statistical thinking