Hi, I am working on a paper using DSEM, computing AR effect and cross-lagged paths. However, the within- person distribution for my participants is characterized by a strong floor effect. I wonder if it is a problem for running DSEM..one possible solution would be to transform my continuous variable in a binary one, but mplus 8.0 (my version) do not allow to run a DSEM for categorical variables at this time. I wonder if there are other possible solutions.
What's your outcome and how many time points do you have? Turning it into a categorical variable (binary or ordinal) is one possible approach. Another is two-part modeling. For both you need to update to version 8.4.