Hacker News Reader

It's a shock when you try to switch mental gears from reading something like a HEP paper, talking about 3 sigma like a big "Well, it could be interesting, probably not." Then you read a front-page "study" about something involving nutrition and "Five people tested, effect confirmed!! Everyone eat Goji Berries!!"

It's mindbogglingly bad, non-science.

View on HN

Depends on the effect size.

You probably only need n = 5 for a confident answer to the question "What happens when you are shot in the head?"

Studies of the form article mentioned helped solidify the link between smoking and cancer also.

People are looking for smaller effects than before, and the incentive to publish any and all significant results is just so great that you get tons of noise.

Maybe, but even then you have the Gabrielle Giffords, Malala Yousefzais, not to mention Phinneas Gages of the world.

Reply on news.ycombinator.com