I have developed three a-priori candidate models that represent factors at three separate spatial scales. A factor in one model is correlated with a factor in another model. (there are debates in the literature about which factor is most influencing the independent variable, and I do not want to draw conclusions about one factor without giving consideration to another.)
But I also learned that I need to create global model that uses all factors from all models. So, is it valid for me to create a global model that has correlated variables?