I am trying to run 2 single mediator models testing two potential mediators without controlling for each other, and then a dual mediator model that tests the same two mediators while controlling for each other. I have a categorical outcome. When I indicate that that the outcome variable is categorical I get identical results for each mediator when I run it in a single mediator compared to when I run it in the dual mediator model--I had expected the results to change given that in the dual mediator model I thought each indirect effect would be controlling for the other mediator. Yet, when I do *not* indicate that the outcome is categorical and treat the ouctome as continuous, I do get different results from the single mediators models compared to the dual mediator model. Can you tell me why my categorical multiple mediator model does not appear to be giving me the indirect effects of each mediator while controlling for the other mediator?
When Mediator #1 is treated as categorical: Single mediator model: B = -.29; CI = -.60, -.01 Dual mediator model (with Mediator #2 in the model): B = -.29, -.60, -.01
When Mediator #1 is treated as continuous: Single mediator model: B = -.19; CI = -.40, -.02 Dual mediator model (with Mediator #2 in the model): B = -.16, CI = -.36, -.02