#### Howdy, Stranger!

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

Supported by

# How to get the baysian factor of simple effect analysis in JASP

Hi,

I have a between factor A, which has 2 levels, and a within factor B, which has 6 levels. I run a Baysian RM ANOVA in JASP, and I found a significant interaction of A*B. Now the question is that I need to analysis the simple effect of the interaction, but how can I get the baysian index of this simple effect analysis?

Thanks

Manuel

• Hi Manuel,

Hmm yes. This should not be too difficult I think, but it has not been done to the best of my knowledge. The thing is that the simple main effects use the information of some of the other cells in the design (this is the benefit over t-tests). I've asked some team members about this as well.

Cheers,

E.J.

• Hi Manuel,

To add to what EJ said, the simple effects analysis is simply a conditional anova, with some correction for multiple comparisons (using the total sum of squares and degrees of freedom to compute the conditional F statistic).

Unfortunately a Bayesian equivalent is missing. What you could do, although this might be tedious if you have many levels, is to manually filter the factorlevel based on which simple effect you want, using the column filtering option and then manually specifying the conditional anova. You would then need to make separate jasp files for each conditional anova, since the column filtering applies to all analyses in a single jasp file. However, this approach has no correction for multiplicity, but it can provide some more insight into the interaction effect (especially if you obtain very strong Bayes factors).

Cheers,

Johnny

• Hi,

Thank you very much EJ!

And thank you very much Johnny! You mean that I need to set 2 jasp files, one for the A1 (one level of the between factor A), and one for A2. A1 and A2 both include the within factor B (six levels). Then, I will run two one-way anova with the within factor B as the independent variable? Am I right?

I am told that this paper, A Bayesian perspective on the Bonferroni adjustment, maybe helpful to calculate the Bayesian factor of the simple effect manually. But I don't know how to do that even I read this paper carefully. Maybe it's easier for you to understand its points.

Thank you both again!

• Hi Manuel

Yes, that is correct!