Hacker News new | ask | show | jobs
by aeternus 3090 days ago
mlloyd has a point, the P value is right on the edge of significance and the study has only 31 people.

Take into account the bias that the negative result (no-correlation) would not pass the bar for publication and you likely have noise.