Effect size calculation in a Bayesian Random-intercept model using brms

Hello,

Is there any consensus about the most appropriate way to calculate an effect size in relation to an interaction term (b*c) from a Bayesian random-intercept model, using brms? The interaction term refers to the interaction between assessment (baseline vs end-of-treatment) and group (placebo vs treatment).

something like this: y ~ a + b*c + (1|id)

Thanks

What do you want your calculated effect size to represent? As in, the (average) difference between what (set of) condition(s) and what other (set of) condition(s)?

1 Like