ABSTRACT
Beyond basic statistics, regression models of various kinds are at the heart of applied statistical methods. Regression models trace how the distribution of a response (or “dependent”) variable-or some key characteristics of that distribution, such as its mean-is related to the values of one or more explanatory (“independent”) variables. Least-squares linear regression is typically introduced in a basic statistics course, while a more general consideration of linear statistical models for normally distributed responses and generalized linear models for non-normally distributed responses is typically the subject of a second course in applied statistics (see, e.g., Fox, 2016, or Weisberg, 2014).
