Hacker News new | ask | show | jobs
by TimPC 4086 days ago
Apparently you can study 13 people and get it blogged about on the ny times. Talk about ridiculous sample size, 13 person studies can say just about anything if the researcher gets lucky because variance is the biggest factor. Interesting random thought -> the number of PhD's who become professors in some fields is well below the margin of error on a study, it would be interesting to see how many people are getting Professorships from having interesting results that eventually fail to replicate. If this number is overwhelming it could be the PhD system in those fields is almost a lottery where having variance trump data is the most likely way to get a professorship.
1 comments

While we might be able to draw stronger conclusions from a larger study, the effects on the participants were strong, and the information might benefit the readers directly. So I'd argue it's worthy of inclusion in the wellness blogs area even if it's not really hard science reporting.