Understading step size (Δ) in Morris Method /jEPlus

edit

I'm applying SA with Morris Method in jEPlus + EA and y trying to understand the theory behind. I would like help to understand the step size Δ for the calculation of Elementary Effects: 〖EE〗_i=(Y(x+∆_i e_i )-Y(x))/∆_i

jEplus documentation steps that Morris takes random step size. I understand then that the step size will depend on each discrete value that I have in my search space. For example Window-to-Wall ratio {0.1,0.2,0.3,0.4}, so my step size (selected by the software) in Morris can be 0.1 or 0.4 for the iteration. Am I correct? What is the case for categorical variables?