forum.smartpls.com

Hello,

I want to ask about moderating effect and multigroup.
E.g.,

I want to examine the moderating effect of B between the relation of A to C. But for B, i have 4 measures where each of them range from 1 ( very negative) to 5 ( very positive).

One suggest me to divide the B to two group, between higher level of B or lower level of B. I did one moderating but only 1 construct so i get the result (i did in other question). Then, how if it is four, is that mean i divide the group of B one by one? To get the info which one belong to higher or lower level??
Or is tha possible to do so?
I desperately need to know

It is my first time to use PLS and happen to be my study need me to do all, reflective, mediating, moderating, multigroup and others


Thank you.

Pls, guys , i need help

It sound like that you should not use multigroup analysis but normal moderator approach. You have four indicators that measure your moderator which is continuous. With that you can use the normal moderating effect approach in SmartPLS.
Splitting a continuous moderator into groups is usually not recommended as you are losing information or building on some idiosyncrasies of the data.

Dear Dr. Jan,

Thank you for your answer.
highly appreciated.


Kind Regards,


M.A