Conceptual modeling of statistical data

A conceptual modeling of statistical objects based on a functional approach is presented. The notion of attribute as a basic modeling construct is used. The presented results show how to deal exactly some very known notions and how to extend and precise them. As a logical consequence of using the typed lambda calculus for description of semantics, some new constructs were proposed. A horizontal and vertical structuralization, horizontal and vertical statistical classification are shown as necessary and useful in modeling of statistical reality.