AN ANALYSIS OF CAUSALITY FOR ROAD ACCIDENT DATA USING GRAPHICAL MODELS

The technique of graphical modelling (Whittaker, 1990) can be used to identify the dependence relationships between variables representing characteristics of recorded road accidents. It allows large multi-dimensional tables to be analysed by looking for conditional independence relationships among the variables. The variables under study can often be divided into groups that are ordered in time or by a hypothesised causal assumption. For these situations graphical chain models (Whittaker, 1990) are used to explore causal relationships between the variables. Some examples are given for a six-dimensional and a ten-dimensional contingency table.

• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/isbn/0080434304
• Supplemental Notes:
  • This book is also available in the USA and Canada from Elsevier Science Inc., P.O. Box 945, Madison Square Station, New York, NY 10160-0757.
• Corporate Authors:

  Elsevier

  The Boulevard, Langford Lane
  Kidlington, Oxford  United Kingdom  OX5 1GB
• Authors:
  • Tunaru, R S
  • Jarrett, D F
• Conference:
  • Third IMA International Conference on Mathematics in Transport Planning and Control
  • Location: Cardiff, United Kingdom
  • Date: 1998-4-1 to 1998-4-3
• Publication Date: 1998

Language

• English

Media Info

• Features: Figures; References;
• Pagination: p. 279-290

Subject/Index Terms

• TRT Terms: Crash data; Variables
• Uncontrolled Terms: Causal relationships; Graphical models
• Subject Areas: Highways; Safety and Human Factors;

Filing Info

• Accession Number: 00763032
• Record Type: Publication
• ISBN: 0080434304
• Files: TRIS
• Created Date: Apr 13 1999 12:00AM