Hello Philo,

For a single outcome variable/dependent variable, Y, the model could look something like this:

Y-estimated = b0 + b1X1 + b2X2 + b3X3 + b4M1 + b5M2 + b6X1*M1 + b7X1*M2 + b8X2*M1 + b9X2*M2 + b10X3*M1 + b11X3*M2

where: X1, X2, X3 are your IVs

M1, M2 are your moderators

Xi*Mj represents an interaction term between IV i and moderator j

and no higher order interactions are hypothesized

For a multivariate outcome, each term (Y-estimated; b0, b1, etc) becomes a vector, having length k, where k is the number of DVs.

Obviously, if you believed that a moderator influenced just one or two of the IVs, then you'd only include the interaction terms for those combinations in your model.

Good luck with your work.

Hello David Morse

one question. I ask to know and understand myself accurately.

Do we call this model Hierarchical Multiple Linear Regression (HMLR)?

No you don't. call it moderation mediation or just interaction model. David Booth

David Eugene Booth

Thank you very much for your answer.

David Morse thanks a lot for this answer, very helpful! One last question: how do I (if neccessary) include the control variables in the equation? Just by simply adding them to the end, like + b13CV1 + b14CV2 etc. with CV = control variable?

interested

Hello again Philo,

Yes, you may add in covariates to the model to reflect the specific research question(s) you're trying to answer.

Good luck with your work.