Hi ,
the two-stage PLS approach for the estimation of moderating effects when formative constructs are involved suggests:
1) Run the main effect PLS path model in order to obtain estimates for the latent variable scores
=> Who do I calculate the latent variable scores (LVS)?
2) Build the interaction term as the element-wise product of the LVS
=> How do I conduct an element-wise multiplication?
Thanks a lot for your help!!
Catharina
Moderating Effects with formative indicators
-
- PLS Expert User
- Posts: 248
- Joined: Sat Jul 25, 2009 1:34 pm
- Real name and title:
Hey Catharina,
ad 1) You can find the latent variable scores after you have run the PLS algorithm in the standard report > PLS > Calculation Results > Latent Variable Scores.
ad 2) Then you have to put your variables (predictor latent variable and moderator latent variable) in an excel sheet. Thereafter you can produce a new variable by multiplication of this two variables case wise. This is your new variable which you can use for the further calculation.
I hope that helps!
Best regards,
Christian
ad 1) You can find the latent variable scores after you have run the PLS algorithm in the standard report > PLS > Calculation Results > Latent Variable Scores.
ad 2) Then you have to put your variables (predictor latent variable and moderator latent variable) in an excel sheet. Thereafter you can produce a new variable by multiplication of this two variables case wise. This is your new variable which you can use for the further calculation.
I hope that helps!
Best regards,
Christian
-
- PLS Junior User
- Posts: 3
- Joined: Sat Oct 23, 2010 5:27 pm
- Real name and title:
HI
please I have the same problem that you are facing; moderation with formative constructs. I did the first step of the main effect model. I have the interaction product of LVS of the two variables but I don't know what to do to put it as a latent variable in the model. I mean hwo to proceed to export the data from excel (LVS product) to smart pls model (as a direct latent variable to the dependent variable). Also I have four moderator reflective LV with two second order factors and one formative predictor composed of two dimensions.
please I have the same problem that you are facing; moderation with formative constructs. I did the first step of the main effect model. I have the interaction product of LVS of the two variables but I don't know what to do to put it as a latent variable in the model. I mean hwo to proceed to export the data from excel (LVS product) to smart pls model (as a direct latent variable to the dependent variable). Also I have four moderator reflective LV with two second order factors and one formative predictor composed of two dimensions.