forum.smartpls.com

Thanks!

anyone? to sum it up, the question is: can I reduce the number of contructs in my model on the basis of correlation analysis, i.e. can I simply leave out exogenous variables, if they show no bivariate correlation to the endogenous variables and take this to drop my theoretical hypothesis that there ...

Hi Christian, thanks for your reply. But I am not sure, if I correctly specified my request. I checked your literature suggestions. Liang et al test 5 control variables. They check, if they do have a significant (direct) influence on the endogenous variables. If I do that, my three controls show no ...

Hi Christian, what should I do if I want to test the moderating effect of a control variable (single measure, continous) on all other connections in my model (some reflective, some formative). Should I include all interaction term into my model at once or better one by one? group analysis is not pre...

Hi there, as fas as I understood, smartPLS gives me the t-values for total effetcs based on the sign change option "no sign change", which I think is the most conservative option (?). But I prefer using individual sign changes as suggested by C. M. Ringle in one of his articles - so how ca...

Hi together, I have a general question for the selection of models: I have a sample of n = 120. My only endogenous construct is "alliance success". From theory I have about 14 potential success drivers and my goal is to test which ones are really relevant for the success. Secondly I want t...

Hi everyone, I also use firm size as a control variable X, measured with the number of employees. Should I transform the data by using ln(X)? I have a distribution with about 80% very small firms and 20% very big firms... Is it necessary to use ln(x)? How to interpret this? What if I want to use X a...

Hi, I dont understand why you would use a latent variable "age"? If it is measured in years (which I assume), there is no need for a measurement model with multiple indicators since it is not abstract but rather concrete with everybody understanding what you mean. You can include "age...

Dear Bido, thank you for your suggestions, well, I maybe could group some of the IV: I am examining joint venture performance and my IV are different critical factors, which I theoretically could group to different stages of the venture or to different aspects like partner-related IV, contract-relat...

Hi @ all, I am still struggling with the question, whether I should apply the PLS-approach or stick to a "regular" linear multiple regression... My model consists of only 1 DV and about 12 IV (I know its a lot for a sample of just n=120). I want to test the direct effects and some interact...

Hi @all,

I have the same problem. In the PLS-Report the value für redundancy of my dependent variable is negative. What exactly does this mean?
Everything else seems to be fine...

best regards,
Carsten

Hello again, I tried your suggestions and thus eliminated all multicollinearity. But I still have some doubts... If I understood you correctly I use PCA with varimax to get VIF = 1 for all variables so that I can a) handle multicollinearity in my structural model and b) handle multicollinearity in m...

Again: Thanks a lot! I will try your suggestions.

best regards,
Carsten

Well, multicollinearity due to too many similar IV actually seems to be the problem as you already suggested. At your opinion: Would it be an adequate solution to aggregate some IV to a 2nd order construct (formativ?) for example by using the LV scores as indicators? With this I could a) improve my ...

Thank you for the reply!

I will check this matter a little deeper. So far I had VIF <2 and KI <3. But as you have elaborated somewhere else in this forum, even VIF slightly over 1 could cause problems.

best regards,
Carsten