Effective altruism, like effective performance in many other fields, relies heavily on empiricism and evidence. Randomized controlled trials are a “gold standard” of evidence which are supposed to be harder to confound. But “harder” doesn't mean “hard”, and many of the randomized controlled trials that shape opinions and policy manage to fall short in one way or another.
In this presentation, doctor and blogger Scott Alexander will review some of the highlights and lowlights of recent research in medicine and public policy with an eye toward how to evaluate and find biases in randomized controlled trial-based research.

Scott Alexander is a psychiatrist and the creator of Slate Star Codex. Slate Star Codex is a blog where topics tend to center vaguely around meta-philosophical ideas of how people evaluate arguments for their beliefs, and especially whether this process is spectacularly broken in a way that may or may not doom us all. In between there’s a lot of cognitive science, psychology, history, politics, medicine, religion, statistics, transhumanism, corny puns, and applied eschatology.