Hello, I am interested in using DSEM to analyze anger ratings between parents and children across an interaction task (30-40 codes for each dyad member). The affect rating is a five-level ordinal variable, although for many of the dyads, people are primarily switching back and forth between two or three of the levels. Existing literature on DSEM suggests that it is compatible with ordinal variables. Is that functionality available in Mplus version 8? I believe that it is not currently possible to create a lagged ordinal variable. Is that accurate?