linear regression makes it possible to predict the behavior of one variable dependent from another independent variable. It has a suction, such as the linearity of the relationship, normality the randomness of the sample and homogeneity of the variances regression doesn’t prove causality.