Modeling and Testing for Control Variables in SmartPLS

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
deeoliv
PLS Junior User
Posts: 2
Joined: Fri Apr 29, 2022 12:38 am
Real name and title: DJ Oliver - Doctoral Student

Modeling and Testing for Control Variables in SmartPLS

Post by deeoliv »

Greetings all,

I'm happy to have this forum as a learning resource. Please forgive me if the answer to my questions is in another post but I could not find it.

1) I'm using SmartPLS to test my model. I have age (what is your age group) as a control variable with the following categories:
  • 18-24
    25-34
    35-44
    45-54
    55-64
    65+
I have age modeled as a categorical variable. I used SPSS to dummy variables for each of the six categories above. In SmartPLS I've created a latent variable for each age category (i.e., 1 for age 18-24, 1 for age 25-34, etc.) and have loaded all 6 latent variables with their corresponding indicators into SPSS. I think I'm doing something wrong because when I run the PLSc algorithm I see very strange values for the path coefficient for the relationship between age and my endogenous variable - the value is -225664878. Can someone please tell me what I'm missing? Should I not load all of the age variables as the same time?

2) I also have education as a categorical variable. My question here is essentially the same. I have 8 education categories and also used SPSS to create dummy variables. Should I also load them all at the same time?

Thanks in advance for your help!

DJ
deeoliv
PLS Junior User
Posts: 2
Joined: Fri Apr 29, 2022 12:38 am
Real name and title: DJ Oliver - Doctoral Student

Re: Modeling and Testing for Control Variables in SmartPLS

Post by deeoliv »

I would appreciate any help on this matter.

Thanks in advance!
Post Reply