>>12758637The 0.05 is just convenience because it's about 2 on a t-test. You can make the critical value something else but it's usually a tradeoff between type 1 and type 2 error. If you're not that worried about type 2 error because type 1 error has much worse consequences you can make the critical value much lower. So it's also context-specific and depends on whether you prefer risking one type of error over another.
