Ordinary Least Squares Regression
- OLS is the standard statistical method of estimating a multiple linear regression model. Its essence is finding parameter values that minimize the sum of squared errors. It involves estimating a straight line (in the case of two variables) in a scatter diagram such that the sum of the squared differences between the observations and the line are minimized. Two-stage least squares, or other more sophisticated techniques, need to be used when systems of equations are interdependent. When the dependent variable is binary, the error term using OLS is heteroskedastic and the method of weighted least squares may be preferred.