Dear all, How would one plot the fitted probabilities from a random effects panel probit model that holds the other variables fixed at their means, while varying two other discrete variables? Best, Andreas