I major in human resource management and am conducting a research about the trickle-down effect of anxiety sentiment. I want to test a 2-2-1 moderated mediation model, with the moderator working in the second stage. My problem is that the moderator I proposed is a level-1 item, which is controversial in the statistical field. Could you please tell me if I can use mplus to test my model, or do I have to revise it to make it feasible? Thank you so much.