I've been writing experimental procedures for a bunch of non-stats folks and for the first time I'm starting to understand the draw of just throwing alpha=0.05 at everything. Getting people to run experimentals *at all* is hard enough, getting them to do it in a non-foolish way feels sisyphean
![cat with tie saying “I hope I [blank] get humiliated” in front a building labeled “humiliation factory”](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:v333fnfm4thmcbyoq3kcolfu/bafkreidmeolcpgpkdztbv5rbg3dl33ujgms6tnbfcrqbeswhorfxxjh35m)
