I ran an error-in-variable model of the form

brm(y ~ me(x,sd_x) + treat)

where x and y are continuous and treat is a factor with 2 levels.

In addition to an intercept and treatment effect, in the Population-Level Effects of the summary table, there is an effect called “mexsd_x”, but I don’t know what its numerical value means. In the example I ran, its value is much smaller than the assumed measurement error sd_x.

Any help would be much appreciated!