forum.smartpls.com

Dear, if there is a significant relationship between A and B (A-->B), and i would like to know what is the right way to test the moderating effect (for example: gender). What test should i use?
some people use (Gender * A --> B) to test moderating effect (SmartPLS 3) and some use MGA to run a t-test between A-->B (male) and A-->B (Female)?
Please help.
Thank you

Generally, both approaches should lead to more or less the same conclusion.
I would use the interaction approach if I have only one path that is affected by gender and the multi-group approach if I assume that several paths in my model are affected by gender.

Thank you, Dr. Becker.
That explanation is exactly what I need: when to use moderator vs. MGA.