Hi,
I have created a PLS-model to show the influence of different organization criteria on the success of a company's sales promotions.
One of the criteria I included is the influence of different departments. Only the three most influencial departments are included in the model. The influence is measured on a 100%-scale and is normalized, so the influence of the three departments considered adds up to 100.
I have so far calculated two models to account for the influence of all three departments. The first model includes department A + B, the second model includes department B + C. All other latent variables are included in both models and are pretty much stable across both models.
I was wondering if I could get rid of one of the models and conclude the influence of the third department (which is not included in the model)?
Any ideas would be welcome...
Stefanie