I’m often irritated by that when a statistical method is explained, such as linear regression, it is often characterized by how it can be calculated rather than by what model is assumed and fitted. A typical example of this is that linear regression is often described as a method that uses ordinary least squares to calculate the best...