Reporting Bayesian ANOVA post hoc tests
Hi,
I've running a series of Bayesian ANOVAs and have to run post hoc tests for some of the factors.
I'm a little confused with what and how I'm supposed to report the output from the post hoc table. Should I report the uncorrected BF10 or the posterior odds and how should I report these?
Would it be reasonable, for example, to say something like:
'Post hoc tests revealed very strong evidence for a difference between condition X and condition Y, BF10 = xyz.'
Can't seem to find much online in this regard so any help would be most appreciated.
Thanks!
Comments
Hi scd21,
A useful thesis on the topic (by one of the members of the JASP team) is here: https://psyarxiv.com/s56mk
I would report the entire table. The uncorrected BFs are just the regular t-tests; the penalty for conducting multiple tests comes in through the prior plausibility, so the t-test Bayes factor is corrected (multiplied) by a prior odds that depends on the number of tests.
Cheers,
E.J.