Hacker News Reader

>Randomization is not the way to select a population who by definition are outliers.

How should the study have been conducted? It is not obvious to me why selecting a random sample is not representative.

I wasn't suggesting they should have conducted the study differently, just that if they sampled businesses generally rather than startups specifically, their conclusions about entrepreneurs don't necessarily apply to startup founders.

Reply on news.ycombinator.com