Hacker News new | ask | show | jobs
by bhickey 2169 days ago
Right. Their confidence intervals overlap: 17.3%-23.0% vs 22.2%-31.0% Factor in multiple hypothesis testing and it's worse than presented.