How to test moderation of multi-answer categorical variable

[GDGOID]

Hello there,

I am still relatively new to SmartPLS and have a quick question about moderators.

I want to show the moderating effects of seniority and length of service moderators in my model.
Likert scale measures were not used in the questionnaire. Instead, respondents selected one option from categories such as: 1 = non-management; 2 = middle management; 3 = senior management in the case of seniority. For length of service, categories included: 1 = 1 to 5 years; 2 = 6 to 10 years; 3 = 11 to 15 Years, etc.

My question: Is it possible to treat these variables as continuous moderator variables and simply run the moderating effect function? If so, could you give guidance or a brief example of how this can be done?

Any advice would be warmly welcomed. Thank you in advance!

[MrcJnk]

I would not recommend using interaction terms for categorical variables as in your case. Instead, you could use MGA to investigate categorical moderator variables. Moderation analysis with interaction terms from my point of view only makes sense for binary variables or ordinal scaled variables from at least 5 point Likert-scales.

Best regards
MJ

[GDGOID]

Thank you, MJ!

[Hengkov]

Use the available approaches for indicators of categories such as OPLSc, Quantile etc.