Accident analysis and its literature may be divided into two groups: detailed technical investigation of accident events and causation at individual locations or involving specific factors; and general surveys of systemwide accident patterns. Trend analysis -- and this paper -- fall into the second category, which is generally oriented to providing reliable data in nontechnical form to civic and government leaders, safety officials and traffic engineers, and other concerned members of the public. Trend interpretation is used both to evaluate the successes and failures of current safety activities and to detect emerging problems as early as possible. This paper takes the opportunity of the inauguration of a new journal devoted to accident analysis to outline some problems in such trend analysis. Perhaps future authors will show how some of the difficulties described might be overcome or, unbeknown to the author, have already been surmounted.
Abstract