Additional Explanation of R2 Results & Control Variables

Frequently asked questions about PLS path modeling.
Post Reply
User avatar
PLS Junior User
Posts: 3
Joined: Wed Feb 17, 2016 8:45 am
Real name and title: Gabrielle Daniels
Location: UK and The Netherlands

Additional Explanation of R2 Results & Control Variables

Post by GDGOID » Sat Jun 11, 2016 12:06 pm

Hello Everyone,

I would appreciate any advice or guidance you could give on the following.
I have a reflective model with 6 exogenous variables which collectively explain c. 33% of the variance in my single endogenous variable (R2 value = 0.325). I was asked recently what could account for or explain the other 0.625. My colleagues suggested introducing a control variable such as tenure or seniority, pointing it to the endogenous variable and the re-checking the effect on the R2 value. That is straightforward!

My control variables have multiple categories however; tenure has 6 for example, and seniority has 3. If I add tenure directly to the model it appears as though it only has a single item. I have the feeling I may need to develop dummy variables/indicators to test this effectively and then point each individual variable to the endogenous variable.

So I have 3 short questions:

1) Is creating dummy variables the correct solution to apply?
2) If so, how do I create dummy variables in SmartPLS 3.2.4 ? I've looked through the forum and elsewhere but can't find guidance on this.
3) Is there an alternative solution that does not include dummy variables?

As always, any help and guidance geatly appreciated!

Post Reply