Interaction term model pulling out interaction effects


I have a brms model with interaction term.
A variable we measured within 2 treatments at 4 different time points.
I am looking at how i could pull out the effect between the treatment and the time points.

The model looks like this:

otm1.1<-brm(formula = bf(DBA ~ Treatment*Assesment+Height+(1|b|Block),
                      hu ~ Treatment*Assesment+Height+  (1|b|Block)),
             data=Analisys, cores=4, file="otm1.1", control=list(adapt_delta=.99))

I want to see the effects of Treatment x Assesment…

Thanks in advance !

ADMIN EDIT: Formatting was added to code by @martinmodrak

I don’t think I understand what you are asking for. If you run summary(otm1.1) you should see coefficients for both the main effects and interaction, but I guess you have done this and it was not enough. Could you elaborate more on what is the question?

Yes i ran summary. But summary gives simply the Treatment x assessment, whilst we would like to see the treatment x assessment 1,treatment x assessment 2,3,4.
We previously ran the DBA, as DBA1,2,3,4 without the assessment, so we looked at the DBA measured at 4 time points… but my supervisors suggest we also check the treatment x time point (repeated measures analysis).
Hope this clarifies it a bit? :/

I am still not completely clear, but maybe the problem is that brms treats Treatment and/or Assesment as continuous predictors, while you would want it to work as a categorical predictor (factor)? I.e. would converting both explicitly to factors before fitting give you what you need?

They are already set as factor :/

I did something similar but in a glmm model, using emmeans… and i pulled out everything i wanted like that… but i cannot figure out how to pull out these interactions…

1 Like

Hi Amelia,

brms is compatible with emmeans, have you tried using that with your model?

1 Like

Actually i haven’t … I didn’t know it is compatible… I guess i can try like that!
Thanks !

1 Like