Tweet from pennmkse

RT @mrtz: 40000 invalid studies or the price of “let’s just assume the noise is Gaussian”: https://t.co/XRvW1vTAvA HT: @fchollet @minzlicht