Factor Score
Posted: Tue Dec 10, 2019 10:40 am
Hi all I understand by clicking the latent variables tab after running PLS algorithm, they are the factor scores of each construct for the data set.
All my scales are 7-point Likert scales from 1 to 7. If the path is positive, by multiplying the factor loadings I would expect positive numbers. However I found something strange. Even for one construct which has only two indicators, all positive loadings, the factor score turns out to have negative. Why is that so? My purpose is very simple, just to show a descriptive statistics of what is the average rating on that construct.
If my model is A and B -> C, while C itself has 3-items, what would the latent variable score of C in Smartpls mean then? A combination of the impact of A and B path weights and C indicators, or just C indicators?
Thanks.
Rebecca
All my scales are 7-point Likert scales from 1 to 7. If the path is positive, by multiplying the factor loadings I would expect positive numbers. However I found something strange. Even for one construct which has only two indicators, all positive loadings, the factor score turns out to have negative. Why is that so? My purpose is very simple, just to show a descriptive statistics of what is the average rating on that construct.
If my model is A and B -> C, while C itself has 3-items, what would the latent variable score of C in Smartpls mean then? A combination of the impact of A and B path weights and C indicators, or just C indicators?
Thanks.
Rebecca