This study improves on the alternative method of obtaining least squares estimates of the parameters of a simple linear regression model by virtue of adding constant to at least one of the associated variables. It is observed that the estimate of the slope is not affected by addition of a constant to any or both of the variables. The estimate of the intercept is a function of the added constant(s). The relationships between the several intercepts of those models involving the transformed values of the variables and that of the original model are then obtained so to facilitate comparison and referencing. These theoretically obtained results have been validated with the help of empirical data analysis.