Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

How to report BF as support for null hypothesis

Hi,

Firstly, I've recently discovered JASP and I have to say that it's the most intuitive stats package I've ever experienced - so a huge thank you for that!

A colleague and I would like to report the BF in a paper, to provide support for the null hypothesis regarding factor 2 (see output below - it's quite small so I attach a copy also). We have shown and reported the lack of significance of factor 2 using NHST, but wanted to use the BayesFactor to support this further. We've seen several ways of reporting BF, but would really appreciate any guidance on how best to go about it.

Apologies for the basic question, as we're new to Bayes stats!

Any help would be appreciated!

Comments

  • Dear Kimberly,

    Sorry for the tardy reply. Looking at the analysis of effects, the data *increase* the prior inclusion probability from 0.263 to 0.750, for an inclusion BF of 3.8; this is mild evidence in favor of Factor 2. So that's not exactly what you were hoping for. Now there is something to be said for using the "across matched models" functionality -- have you checked that out? I will ask one of my students to send you a link to a preprint that we've submitted with more information on this.


    Cheers,

    E.J.

Sign In or Register to comment.