DATA ON 1008 QUEBEC ROAD VICTIMS OBTAINED FROM LINKAGE OF GOVERNMENTAL RECORDS HAVE LED TO DIFFERENT TYPES OF ANALYSES. THIS ARTICLE PRESENTS AN APPLICATION OF CORRESPONDENCE ANALYSIS TO THESE DATA. WE HAVE CHOSEN ONE EXAMPLE WITH ITS SPECIFIC RESULTS SHOWING EXPECTED ASSOCIATIONS BETWEEN VARIABLES AND NEW, UNEXPECTED ASSOCIATIONS WHICH ARE PLAUSIBLE. THE RESULTING PATTERNS INDICATE A CLEAR DISTINCTION BETWEEN HOSPITALIZED AND NON-HOSPITALIZED VICTIMS AND BETWEEN A RURAL AND AN URBAN TYPE OF ROAD CRASH. THIS MULTIDIMENSIONAL APPROACH GIVES AN OVERALL PICTURE OF THE PATTERNS AMONG THE VARIABLES STUDIED, INDICATING WHICH SUBSETS OF VARIABLES MIGHT BE INVESTIGATED FURTHER.(Author/publisher).
Abstract