ABSTRACT
Functional models can be constructed for any kind of object or situation as far as their properties can be represented by variables. An interesting distinction arises if one considers a collection of individual units, say = {ω1, . . . ,ωn}. Functional models may then refer either to individual members of , or to as a set. In the former case, models connect variables characterizing individual members of ; such models will therefore be called individual-level models. In the latter case models connect statistical variables or distributions that characterize as a set of individual units; such models will be called population-level models.
