Essentials of Multiple Linear Regression

This chapter describes extensions of regression analysis that accommodate multiple predictor variables. Although many applications of multiple regression analysis are similar to those of simple regression, other applications are considerably more complex. The chapter begins with a gradual introduction to the basics of multiple regression analysis in the context of the two-predictor model. Multiple regression analysis involves the simultaneous use of two or more variables in predicting an outcome variable. Partial correlation measures the degree of linear relationship between two variables after statistically controlling for one or more other variables. Ordinal regression is available for studies that use a dependent variable that consists of ordered categories, and tobit regression may be useful when many of the dependent variable scores fall at the minimum or maximum points of the distribution. Controlled Vocabulary Terms coefficient of correlation; Multiple regression; regression coefficient