Multiple regression models for quantitative and categorical data
Multiple regression is a statistical tool used to determine the association of multiple variables on an outcome. This chapter provides examples of model building with quantitative and categorical data. Categorical data are variables that are mutually exclusive such as ethnicity, religion, or gender. For continuous variables participants may vary in terms of amount, whereas for categorical variables participants vary in type. The chapter, in addition to examining quantitative and categorical data, also introduces curvilinear regression model development. It focuses on the more common powered vectors regression as it has been typically used in the exercise science literature. Polynomial regression models are an extension of linear models. Logistic regression is the appropriate technique to use when outcome measures are dichotomous or multinomial. Statistical regression consists of three different analytic methods. They are the forward, backward, and stepwise models.