Hi all,
I have two questions concerning the theoretical background:
Please consider three endogenous LV in my model of six LV in total:
SE (reflective)
SS (formative)
ES (formative) with the relationships SS-->SE and ES-->SE.
1) When I check for discriminance validity of the reflective LV SE I look at the LV correlations and ensure that Fornell-Larcker-criterion holds. In my case it does not (SE highly correlates with SS and ES). But this correlations makes sense in theoretical terms. How do I proceed?
2) Is this high correlation the cause for very low R2 of SS and SE but for extremely high R2 of SE and if so why?
If not - do I have to reject the model or is a high R2 of SE sufficient?
Thanks in advance to everyone. Every comment is highly appreciated.
Cheers, Lennart
Relationship of LV correlation and R2
-
- PLS Junior User
- Posts: 4
- Joined: Tue Nov 28, 2006 10:08 am
- Real name and title:
-
- PLS Expert User
- Posts: 139
- Joined: Wed Jul 05, 2006 1:43 pm
- Real name and title:
- Location: Coimbatore, India
discriminant validity
You need to check if the RMS values of Ave values of the constructs is more than construct correlations. If not, it implies that the two constructs are one construct and not two.
R2 value has no role in this.
R2 value has no role in this.