Chapter 1 introduces readers to the elementary concepts of the General Linear Model (GLM). The GLM is the most widely used model in applied statistics. It states that an observed variable can be explained from a number of predictors that each carry their individual weight. The weights are estimated from the data. The predictors are either metric or categorical. The error terms quantify model-data discrepancies and can be conceptualized as a convolution of unconsidered factors that impact the dependent variable, model imperfections, and measurement error of variables. The model is linear in the model parameters because none of the weights, that is, none of the model parameters is raised to a power different than 1. The model can, however, be non-linear in the independent variables (e.g., raising continuous predictors to a power different than 1 results in polynomial curvilinear regression)

Two different classes of variables capable of affecting the intensity of emotions are introduced: global variables, which can influence the intensity of many different emotion types across different emotion groups, and local variables, which have relatively local effects, influencing only the intensity of particular emotions or emotion groups. Examples of global variables are presented, including sense of reality, psychological proximity, unexpectedness, and arousal, along with examples of local variables such as deservingness, relevant for Fortunes-of-others emotions such as Schadenfreude, and likelihood, relevant for emotions involving envisaged events. A detailed discussion of examples of emotion types involving both kinds of intensity variables is provided. Also discussed is the relation between global and local variables on the one hand and the central variables discussed in Chapter 3 on the other, with particular attention paid to the issue of how such interactions contribute to the overall intensity of particular emotion types. The difficulty of calibrating intensity across different emotion types is discussed.