# Interaction of two responses in multivariate model in brms

**URL:** <https://discourse.mc-stan.org/t/interaction-of-two-responses-in-multivariate-model-in-brms/13791>\
**Category:** brms\
**Created:** [March 21, 2020, 8:12pm UTC](https://discourse.mc-stan.org/t/interaction-of-two-responses-in-multivariate-model-in-brms/13791 "2020-03-21T20:12:16Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Benedikt\_Holtmann](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/benedikt_holtmann/32/7465_2.png) [@Benedikt\_Holtmann](https://discourse.mc-stan.org/u/Benedikt_Holtmann)\
**Post date:** [March 21, 2020, 8:12pm UTC](https://discourse.mc-stan.org/t/interaction-of-two-responses-in-multivariate-model-in-brms/13791/1 "2020-03-21T20:12:16Z")

</div>

Hi,

I just would like to know whether there is a way to fit an interaction term between two response variables in a multivariate model in brms?

For example:

model\_1 \<- brm(  
mvbind(y1, y2, **y1:y2** ) ~ 1 + x1 + x2 + (1 |p| g1) + (1 |q| g2),  
family = “gaussian”, chains = 4, cores = 4,  
iter = 2000, warmup = 500, thin = 1  
)

Thank you,  
Benedikt

---

<div class="post-metadata">

**Author:** ![paul.buerkner](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.mc-stan.org/paul.buerkner/32/3303_2.png) [@paul.buerkner](https://discourse.mc-stan.org/u/paul.buerkner)\
**Post date:** [March 21, 2020, 9:38pm UTC](https://discourse.mc-stan.org/t/interaction-of-two-responses-in-multivariate-model-in-brms/13791/2 "2020-03-21T21:38:33Z")

</div>

I am not sure this makes conceptual sense. Why would you like to fit such a model?