Moderation effects for different samples

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
Soeren Baersch
PLS Junior User
Posts: 1
Joined: Fri Jun 12, 2020 8:09 pm
Real name and title: Soeren Baersch

Moderation effects for different samples

Post by Soeren Baersch »

Dear Researchers and professors,

For a current research model, I collected data from German and Chinese consumers regarding Mobile Payment Intention to use. My research framework: extended UTAUT Model.
I already tested for the German and Chinese sample the age, gender or shopping frequency (high vs low) within the sample to obtain invariance with a MICOM procedure.

Now, I would like to test if there are any significant differences between Chinese and German consumers. But before, I can calculate the p-value for the difference between the path coefficients and moderators. I need to know if the effects are driven by the coefficients or variance between both samples. I want to use the suggested method by Dabholkar and Bagozzi 2002 (DOI:10.1177/0092070302303001) (e.g., Im et al. 2011; DOI:10.1016/j.im.2010.09.001 use this method) to test this.

I´m not sure how to do this! Especially with the four Models. Should I merge the dataset and test for the nationality (German & Chinese) as a group or not? Can I use this method with SmartPls 3?

Or is there any other method that fits better?
Thank you for your support!

Best Regards
Sören Bärsch
Post Reply