Predictor variables for random slopes

[Keith Lau]

Dear Phil,

I am asking for help about adding predictor variables to random slope for multilevel IRT models.

For random part, I used "random = ~ 1 + prior_score|group". It works well with the individual-level predictor " prior_score " and grouping variable "group".

Now I want to add a school (group) level predictor "school_type" to the random slope of " prior_score ". I looked up the mirt help and examples, but still cannot figure out how to modify the code.

Your help is most appreciated. Thanks a lot.

CW