Although regression for environmental data is more commonly used to analyze intervariable relationships where both the response (i.e., *Y*) and explanatory (i.e., *X*) variables are *quantitative* (*continuous*) variables, *qualitative* (*categorical*) variables can also be analyzed using regression methods. The various types of variables used in data analysis are described in Section 16.2. A main difference between continuous and categorical data regression is that the former involves only numerical quantities, which can therefore be directly used for the regression analysis, while the latter often involves nonnumeric qualitative data, in which case a secondary coding system comprising artificial *dummy* or *indicator* variables is required for the data to be usable for regression (see Section 16.3). Categorical data regression can be classified into two broad categories, namely, (1) the response or *Y* variable is the usual quantitative and continuous data type, while the predictor or *X* variables are either all categorical or a combination of categorical and quantitative data types; and (2) the *Y* variable is categorical (i.e., qualitative), while the *X* variables can be all categorical, all quantitative, or a combination of categorical and quantitative data types.

For the first type of categorical data regression, the models are often described as analysis of variance (ANOVA) or analysis of covariance (ANCOVA) models because ...

Start Free Trial

No credit card required