Clarification on conditional_effects(..., robust = TRUE)

Interfaces brms
1

@paul.buerkner and others …

Seeking clarification that while robust = TRUE in the code below calls for the median, it is nonetheless, overridden by posterior_epred that provides only the mean — despite the robust option.

If so, then maybe a notice to that effect in the documentation for conditional_effects(...) might be warranted?

conditional_effects(..., method="posterior_epred",  ... , robust = TRUE)
2

These two things operate on different levels. robust says something about the kind of summary statistics that are computed over the (implied) posterior of a chosen quantity of interest, while posterior_epred tells you which quantity of interest we look at, that is in this case, the mean of the posterior predictive distribution.