Marginal effects #

Summaries of the association between a change in a regressor and a change in the response variable.

Syntax #

drivers: true

Usage #

When drivers: true, the model API creates additional JAGS objects, which are used to evaluate the marginal effects of explanatory variables in the model. By default, it shows the effect of each variable over its empirical range, holding all other variables at zero (the mean of scaled continuous variables or the reference level of categorical variables).

Note

The behavior of drivers: true depends on the type of variable. The effects of continuous variables are shown on their original scale. The effect of indicator variables is shown on a 0-1 scale. Interaction terms are (perhaps unhelpfully) left in scaled form, and require additional work to plot properly. See this write-up for more!

There is an additional, optional statement called crossings that can be called alongside drivers to evaluate different combinations of explanatory variables. For instance, if we are interested in the effect of \(x_1\) not just at the mean of \(x_2\) (i.e., when \(x_2=0\) ), but when \(x_2=\mathrm{min}(x_2)\) or \(x_2=\mathrm{max}(x_2)\) , we would specify:

drivers: true
crossings:
    - min_seen
    - zero
    - max_seen

Warning

We do not recommend using crossings unless that specific line of inference is needed, as it can drastically increase model run times and the size of the model object.