Hypothesis testing for interactions in a multivariate model

Interfaces brms
1

Hi all,

Hoping to find so clarity!

I am running a brms multivariate model as described below:

model_x_formula <- bf(mvbind(risk1,
                             risk2,
                             risk3,
                             risk4) ~  
                              Condition * Gender + BMI + (1|ID))

I have main effect hypotheses, which I can check with no issue, but I would also like to better understand the interaction between condition*gender. Condition has level A and B, and gender Male and Female. More specifically, I would like to see if the condition impacts men differently than women, and see the direction of the change. By checking related topics on the forum, I came to this formula for hypothesis testing:

hypothesis(model_x, “risk1_ConditionA:GenderMale > 0”)
hypothesis(model_x, “risk1_ConditionA + risk1_ConditionA:GenderMale * -1 < 0”)

My understanding is that the first hypothesis tells me the effect of the interaction between condition and the default level of gender, while the second will tell me the effect of the interaction between condition and the other level of gender, female. Is this correct?

Looking forward to any feedback to understand this better and thank you in advance!