Hacker News new | ask | show | jobs
by aquafox 591 days ago
You shouldn't publish based on p < 0.05, but on orthogonal data you gathered on the instances that showed p < 0.05.