Hi all
I wan't to use unstandardized path coefficients as they are easier to interpret (1 scale unit up in eksogen latent, equals 1 scale unit up in endogen latent var). All my data are on the same scale, so the different scale issue is not a problem here.
I've corrected the standardized coefficients and ends up with "index values -> Path Coefficients" if I use loadings to correct the coefficients, but wouldn't it be more right to use the outer weights?
Furthermore this could result in a path coefficient above 1 but I have never seen this in a published article and this makes me a bit insecure about the approach (a lot of articles claims, to have unstandardized path coefficients especially the ECSI articles)
One other thing is the use of "Original" or "Mean0, Var 1" data metric. When is it appropriate to use one over the other? I cannot seem to figure out the connection between the results from the different options.
Thanks in advance
BR
Sten A
Unstandardized path coefficients (calculate and interpret)
-
- PLS Junior User
- Posts: 1
- Joined: Thu Dec 14, 2006 5:19 pm
- Real name and title: