So I am testing my own model but it is more complex than I have dealt with before. I know I need hierarchical regression, and I'm using SPSS (I have no choice in programme, sorry), but I can't find a tutorial close enough to what I'm doing to be sure it's right.

I have one IV (3 categories but dummy coded into 2 variables) and one continuous DV. Between these two I have two continuous mediators, M1 and M2. Three of the four relationships between these variables have moderators acting upon them (one on IV > M1 relationship, one on M1>DV relationship, and one on M2>DV relationship).

None of my extra demographics appeared to correlate with my DV in initial correlations, so I have no control variables to enter in step 1.

My first attempt I entered the two dummy coded variables of my IV into step 1, with the DV in its place.
Step 2 - both mediators.
Step 3 - all 3 moderators.

I have no idea if this is the right process and would love any feedback or corrections. Should I be treating each mediator in separate regressions even though they have the same IV and DV? I know that would damage my stat power so would love to keep the number of tests minimal.

Hope that makes sense and thanks in advance for your help