j is interpreted as the average effect on Y of a one-unit increase in Xj, holding all other predictors fixed (Gareth et al. 2014). Using population and precipitation as regressors gives us: