Multiple regression models

In regression models there is a clear separation between the regressed variable and the regressors (explanatory variables):

This does not necessarily mean that there is a causal relationship, but it is enough to classify regression models as dependence models. Regression models arise naturally for dealing with metric variables, but we may use binary variables to model qualitative features in a limited way. We may use regression models for confirming a theory, by testing the significance of individual coefficients or the overall significance of the whole model. However, we may also use the technology as an exploratory tool, by running a sequence of regression models involving different sets of regressors. Furthermore, logistic regression models allow for a qualitative regressed variable taking only two values. Alternative methods, such as discrete discriminant analysis, may be used for the case of a dependent categorical variable assuming more than two values.