Control Variables - NOT moderators

[sonja.h.ms]

Hi there!

I am wondering if it is possible to integrate control variables in the SEM model in SmartPLS 3, and if yes, how this can be done?

Specifically: I want to analyze the influence of a certain message in an advertisement (dichotomous: message integrated = 1, not integrated = 0) on the attitude toward the product and purchase intention. Thereby, attitude is also expected to influence purchase intention. I also have a moderator on these effects which is product knowledge, but I furthermore want to control for biasing characteristics such as product involvement.

I know how to integrate moderating effects, but how is it done if I only want to control for biasing effects, i.e., make sure that product knowledge does not influence the manipulation effects?

Thank you so much for your help!