My model consists of one binary independent variable (IV), one multi-categorical moderator (MOD) and a continuous dependent variable (DV). 
IV and MOD are highly correlated (if IV=0, no moderation, if IV=1, MOD=1, 2 or 3) on purpose.
When running the Hayes Process macro in IBM SPSS 28, the following information including several errors is presented, 

"OUTCOME VARIABLE:
 DV

SINGULAR OR NEAR SINGULAR DATA MATRIX.
Error encountered in source line # 49489

Error # 12492
An attempt has been made to use previously undefined matrix (or scalar).
Execution of this command stops.
Matrix - 'B' is undefined
Error encountered in source line # 49489

Error # 12322
Right hand side of COMPUTE is undefined.

*********************** ANALYSIS NOTES AND ERRORS ************************

ERROR: A linear or near linear dependency (singularity) exists in the data."

I have created dummies for the moderator with same result.

Your help or suggestions are appreciated.

Hello Jon,

Thanks for your reponse. I ran the linear regressions successfully. However, it's my understanding that the Hayes Process macro offers significant advantages over simple linear regression for moderation analysis and it can enhance the reliability of the moderation analysis results. I've checked various sources and for my model, the macro should work. I was hoping to learn how to get it to work.

Thanks again. Regards, Twan Oosterveld

Hi Jon,

I am using SPSS 29 and have a similar error as well. When I use model 4, everything is working (X, Y and M - continuous variables) but when I use Moderator (binary) in model 58/59 I get the error message mentioned above. I tried running a simple moderation analysis using model 1 and still the same error.

I tried looking for collinearity as you suggested and my VIF values are <1.1, which is not high from my understanding.

Thanks,

Saahithh