At the TRRL/SWOV Workshop on Accident Analysis Methodology, held during the International Symposium on Traffic Safety Theory at Amsterdam in 1988, the need for the establishment of a methodology for the analysis of road accidents was firmly stated by all participants. Data from many countries cannot be compared because of lack of unification in data collection and data handling. Furthermore, the comparison of the results from different analyses on these data are obscured by the use of different techniques for analysis. A wide variety of techniques has been applied, such as ordinary (Multiple) Linear Regression Analysis (MLR), Log-Linear analysis (LLA), Generalized Linear ractive Modelling (GLIM), primarily applied at the Transport and Road Research Laboratory (TRRL) in England and techniques used at the Institute for Road Safety Research (SWOV) in the Netherlands, that are classified here as Qualitative Data Analysis (QDA) (including canals, a generalized canonical correlation program). Research is being carried out to compare the results of a number of analyses with QDA techniques applied to data from TRRL, with the outcomes of the GLIM analyses that were already carried out at TRRL. TRRL, SWOV and the Transport Studies Group (TSG) of University College London were the major parties in the discussion on the methodology of road accident analysis briefly reported.
Abstract