forum.smartpls.com

Yes it is true. In SmartPLS 3 you get the collinearity statistics for the inner and outer model. No need for SPSS anymore. You interpret the VIF like you did in SPSS.

Yes, if your interaction term is significant then you can conclude a moderation effect. The direct effect of the moderator must be included in the model to get correct parameter estimates, but it must not be significant.

I think the procedure sounds reasonable. With the SmartPLS 3 you get the unstandardized LV scores in the IPMA procedure. Moreovere, you will get there the unstandardized coefficients for the inner structural model and also means of the LV scores, so I think there is no need anymore to go to SPSS to ...

It only considers the direct effect. What the effect size f² calculates is the change in R² if you exclude a variable from the model. If you exclude your exogenous variable, the mediators will stay in the model and will still effect the final dependent variable, hence only the direct effect of the e...

Can I download the new version without losing all my models/calculations?

Of course.

Do you use the latest version 3.1.6? In an earlier version there was a bug regarding the indirect effect calculation. If you have the latest version, then you somehow modeled an indirect relation between A and C (maybe through B?). Otherwise, there would be no result for the indirect effect of A-->C...

Simply because they are not the same (see above).

Model:
A --> B --> C and A-->C

Direct Effect: A-->C
Indirect Effect: A-->B * B-->C
Total Effect: A-->C + A-->B * B-->C

If your indirect is the same as the total effect, then you don't have a direct effect in your model.

There should be no difference. Usually, people who report this problem used wrong data settings. For example, it can make a huge difference if you used the wrong number format (i.e., decimal sign; dot or comma).

Of course the results should be equal. All of the similar reported problems have been data settings problems. Did you have missing values that you specify correctly? Did you check whether the numbers that are imported are really the same? (check the data used for calculation output). Finally, you co...

To run the MGA you need to define a grouping variable for the dataset (in the data view). For example, you could use a dummy variable for that purpose. However, you could also use that dummy variable in your model and let that dummy predict your dependent variables that you think are influenced by y...

I think the path coefficients are the closest to what you are looking for. It reports the regression results between the latent variables scores in the inner (structural) model.

I have never heard of a "Complete Function" in our software and I have to know it. There is a PLSc, which is a consistent version of PLS, but this won’t help as it is based on a normal PLS run and corrects the results in a certain way. Yet, the singular matrix is not necessarily only a pro...

You should definitely use your sample size (N) for the number of cases. In SmartPLS 3 we have removed that setting and derive the number automatically to avoid user errors. Generally, t-values are affected by sample size. The higher the sample size the higher the statistical power and hence the t-va...

You will find the Cross-Loadings in the Discriminant Validity section