Hello,

I have used the PROCESS macro several times and have had success except now. Have a model (Hayes #4) with 8 mediators. The predictor and mediators are continuous while the outcome is binary. PROCESS estimates all the paths, but fails to generate the indirect, direct, and total effects. So my question is, can I calculate these myself? The paths to the mediators are continuous estimates while the mediator and predictor are log odds - two very different metrics. I guess I am asking is there are way to make these a common metric and then use Hayes' equations to calculate the effects? Or is that just not reasonable?