I am using Smart PLS the first time and try to find out how to use the programm properly. Although I run the introduction videos on this page and read lots of commons in the forum there are still lots of questions left:
1.) When do I use original data and when Mean 0, Var 1? (My original data from spss are not standardized. I tried both and there are obviously big differences, especially in my one reflective LV.)
2.) After running the PLS Algorithmus I had to recognise that some weights are negativ or above 1. Is that by all means possible or do I something wrong?
3.) What is the relevance of the index value in the model output? Do I need these information to interprete the results?
4.) Still to the output: When do I use standardized values and when unstandardized?
5.) I found different statements for testing the validity of formative LV. Some authors say, that one can only check the multicollinearity and others say you also should have a look on the weights and their significance. What is right or best practice?
6.) I would like to test a moderating effect between all my exogenous LV and my one endogenous LV (e.g. the influence of a kind of sport on the relationship quality of a sponsorship). Is this possible?
I know these are a lot of questions. But hopefully somebody can answer me some of them.
Thanks in advance. Andrea
beginners questions
-
- PLS Junior User
- Posts: 4
- Joined: Thu Oct 23, 2008 3:21 pm
- Real name and title: