Hacker News new | comments | show | ask | jobs | submit login

Not even a large sample will help against systematic errors.



Good point; upvoted. However I was just trying to address the idea that there are ways of minimizing problems in study design. No study is ever perfect, but many of them are sufficient.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: