forum.smartpls.com

The only possible option is to exclude the variable from your group model. This should not be a problem because you already control for the variable by grouping your data.

First, you need to think which is the correct test for your example. Do you want to test mean differences? Or what do you want to test?

Please also use the search function of the forum. It is a very useful information source. I think this topic has been covered already.

That means that most of the fit measures are not defined for your model. This likely happens when you have repeated indicators. Having those in your model means that you have a correlation matrix with correlation between variables (the repeated indicators) that are 1, which means perfect collinearit...

Sure. That is a method specific feature and not a software feature.

Well, yes based on the structural model that you show 10 is the largest number of predictors that point to a construct (HOC_e). However, if you model the HOC with the repeated indicator approach (which seems to be the case in your model) and you have a formative HOC and use Mode B to with the repeat...

It is certainly possible, but unusual. How do you assess significance? Based on t and p-value? For the indirect effects it is usually better to look at the confidence intervals as the bootstrap distribution might be non-normal and therefore using the t-statistic might not be appropriate. You can als...

That sounds very strange. Effects sizes of moderating effects are usually quite small. Do you use the f² effect size from the SmartPLS output or a self-calculated one?

You need to import them without the % sign. Whether you use 5 or 0.05 should not make a difference as long as all values are on the same scale.

If you have such small sample sizes for the group results are certainly not very reliable. The biggest chance is that you are missing out important differences because of low statistical power. However, if your groups are not very representative for the general population for which they stand, you m...

If you have only one predictor and you exclude the predictor you explain/predict nothing. It is a bit of a debate whether it makes sense to calculate the indices for such a variable with R² excluded etc. being 0 or whether they are not defined.

I would say that for f² is makes sense, but not for q².

You should copy the undstandardized latent variable scores from the IPMA (IPMA > Latent Variables > Unstandardized) into SPSS, Excel or some other software that can run normal regression analysis. You should get the same results for the path coefficients as in (IPMA > Path Coefficients > Unstandardi...

You just need to also include the group information for the second-stage model. The observations in the first-stage model are in the same order as in your data set. Hence, you can just include the grouping variable from your first-stage model in your second-stage model.

Please send your model and data to the support. They can help. http://support.smartpls.com/

Generally, you should make sure to use the latest version (version 3.2.7). You should also state which algorithm you use (PLS or PLSc) and which bootstrap settings you use (basic or complete), etc.

SEM in general is correlational. Some people use causal language because they believe their theory is strong enough to make causal claims based on correlational data. Basically, you can make the same claims that you would do based on a multivariate regression; just with multiple-indicator measuremen...