Use of R Squared value.

**How close the data are to the fitted regression line**

**R-Squared** is a statistical measure of fit that indicates how much variation of a dependent variable is explained by the independent variable/variables in a regression model.

**R-squared** should accurately reflect the percentage of the dependent variable variation that the linear model explains.

**R-Square** does well in simple linear regression (with 1 predictor) but for multiple regression you may need to have an adjusted R-Square calculated because as you add more predictors in your model, you R-Square tends to increase (always) and hence in case of multiple predictors always go for adjusted R Square not R-Square.