Maybe all studies should be preregistered, including their methods... like this one was?

https://osf.io/vzdh6/