I have a SEM with a dichotomous DV, 3 latent IVs and several covariates. I am using WLSMV. I wanted to see probability curves. I have run plot 2. Then I viewed the graph and chose:
1) item characteristic curves 2) for a single item 3) chose my DV 4) for all categories 5) choose the IV of interest
Since the plot holds all of the other variables at their means, I would anticipate that I could run this separately for each of my 3 latent IVs and at "0" the probability should be the same--since all variables are at their means.
However, I see slight differences at "0." Why is this?