Predicting a quantittive outcome from 2+ predictior variables while controlling for potential confounding-covariate variables.