Daniel LakensDec 19, 2023
This tweet by @alessandro_nai is one of the real life examples of deviating from a preregistration that I use in my new preprint. https://osf.io/preprints/psyarxiv/ha29k Mistakes in a preregistration happen! If they are obvious enough, peers should agree they do not impact the test severity.
OSF

Show threadLeonie Sieger
@lakens Confession: I wrote in a prereg that I'll do an ANOVA and later realized that of course I have ordinal data (assuming ad-hoc item likert scales are ordinal) and should do a friedman test instead. Then I learned that the friedman test is called friedman ANOVA by some, so... somehow fine?
(Anyway, it's a computer science paper, so it's already weird that I preregistered. Or discuss my choice of methods in it. Or study people.)
Dec 19, 2023 at 1:33pmWeb
Show threadDaniel LakensDec 19, 2023
@leoniensieger ha! Thanks for sharing. I think this is a good example of an error (it is even is one) that has little consequences :)