Conceptual Data Structures and the Statistical Modelling of Road Accidents

Abstract

The aim of this paper is to review the structure of road accident data, the database framework within which it is stored, and the potential for exploiting a hierarchical structure using multilevel statistical models. Most national accident databases regard accidents as the primary units of observation, with other characteristics stored as attributes. But it is more natural to picture the network, accidents, and other variables as a collection of related objects within a hierarchical system, which can be achieved using object-oriented database technology within a Geographical Information System (GIS) framework. This would permit more efficient data capture and storage, facilitate analysis of accident frequencies as a function of road layout, and facilitate the development of multilevel statistical models.

Citation

Lupton, K., Wing, M. and Wright, C. (1998), "Conceptual Data Structures and the Statistical Modelling of Road Accidents", Griffiths, J.D. (Ed.) Mathematics in Transport Planning and Control, Emerald Group Publishing Limited, Leeds, pp. 267-277. https://doi.org/10.1108/9780585474182-026