Bivariate Regression

Bivariate regression: A regression analysis that uses one independent and one dependent variable.

Multiple regression: A regression analysis that uses two or more independent variables and one dependent variable.

Ordinary least squares (OLS) regression: A common procedure for estimating regression equations that minimizes the errors in predicting the dependent variable.

One Independent Variable and One Dependent Variable: Bivariate Regression

Bivariate regression is an extension of bivariate correlations, so we will begin the discussion of OLS using some of the data. These data were from a sample drawn from the Juvenile Defendants in Criminal Courts. They measure the number of charges filed against each defendant, and the dollar amount of the fine imposed. (This sample has been narrowed to defendants who were convicted and penalized with fines.) The data from Table 13.4 are reproduced.They are sorted in ascending order first according to the charges variable (the IV). Sorting them this way prepares the data for easy plotting on a graph.