Independent latent variable can get R-square 0?
- jakesbae
- PLS Junior User
- Posts: 1
- Joined: Wed May 26, 2010 4:17 pm
- Real name and title:
- Location: Republic of Korea (South)
- Contact:
Independent latent variable can get R-square 0?
(trust -> four organization cultures -> lie)
(this is corelations of trust factors.)
I was executed analysis by PLS algorithm with case wise replacement and path weighting scheme.
But, I can get 0 value in Independent latent variable. (first left circle)
my professor was said "Humm.. It's R-squre value. this value's range is 0 to 1. why you get 0? I think it may have some ploblems. please check something is going wrong." why I get 0 in a independent latent variable?
and I'm going right way what analysis this path diagram?
please give me some advice to me.
(n=84, It has some missing values)
Jake S. Bae (Jake Seongho Bae)
Industrial and Organizational Psychology Student
jakesbae@orblab.kr / +82505077587
Industrial and Organizational Psychology Student
jakesbae@orblab.kr / +82505077587
-
- PLS Expert User
- Posts: 37
- Joined: Tue Mar 25, 2008 9:16 am
- Real name and title:
Hi jakesbae,
Of course your IV has an R² of 0, that is why it is INDEPENDENT, which means that it is not explained by another variable (if it was a dependent variable, some of the variance [measured by R²] could be explained by another variable). I wonder why your professor is not familiar with this.
So this is not a problem, but looking at your measurement model, you have some other issues here. The negative loadings in the mediating construct could be an indicator for multicollinearity which you might want to assess.
Best regards, Stephan
Of course your IV has an R² of 0, that is why it is INDEPENDENT, which means that it is not explained by another variable (if it was a dependent variable, some of the variance [measured by R²] could be explained by another variable). I wonder why your professor is not familiar with this.
So this is not a problem, but looking at your measurement model, you have some other issues here. The negative loadings in the mediating construct could be an indicator for multicollinearity which you might want to assess.
Best regards, Stephan