For example, linear regression could be used to model the dependence of a clinical response on the dose of medication. However, the slope and intersect parameters of this regression may be different for different groups of patients: elderly patients, e.g., may show a stronger reaction to the medication, so that the slope of their regression line would need to be steeper than that of younger patients. In this example, the model of interest is the regression between dose of medication and clinical response – however, the model parameters should be chosen differently in the two (or more) groups defined by the covariate age. Another example and visualization are given in the section “Further application examples”.
