Closed jpoehl closed 4 months ago
https://github.com/dmphillippo/multinma/blob/dba35ad233b342290e08aa84a0e79ce49be8ae17/vignettes/example_dietary_fat.Rmd#L101-L102
The description of the RE model specifies $\mathrm{N}(0, 100^2)$ as the prior distribution for treatment effects and intercepts, but the code uses 10, not 100, as its scale argument (unlike, for example, the FE model above).
Well spotted, thanks!
dmphillippo
commented
4 months ago
https://github.com/dmphillippo/multinma/blob/dba35ad233b342290e08aa84a0e79ce49be8ae17/vignettes/example_dietary_fat.Rmd#L101-L102
The description of the RE model specifies $\mathrm{N}(0, 100^2)$ as the prior distribution for treatment effects and intercepts, but the code uses 10, not 100, as its scale argument (unlike, for example, the FE model above).