Hacker News Reader

if you are p testing this isn’t the case. A positive result is a much stronger assertion

Sure and of course you shouldn't change your hypotheses. That's kind of the whole purpose of pre-registrations at a meta-level of things.