A survey conducted on a young adult driver population (in Quebec) in 1984, using accident records compiled over a period of 6 years (1977-83) generated an impressive number of highly valid variables. Jessor's conceptual framework was retained for its great capacity to predict (and explain) problem behaviours among youth. Psychosocial studies have not been too successful in the past in predicting alcohol-induced car accidents. But Jessor's team never applied their framework to this neglected topic. So we attempted to fill the gap and did use it as a test framework. To complement this, human ecology concepts from William Ogden's work were borrowed to enlarge the scope of our analytic model and encompass the most relevant ecological factors. Using a 30-page confidential questionnaire, 938 individuals throughout Quebec were contacted by mail after being systematically sampled from an exhaustive list of 10534 names representing the whole target-population. The questionnaire was returned, completed, by 515 individuals, that is 53% of the sample. La Regie d'Assurance Automobile du Quebec (Quebec Automobile Insurance Agency) forwarded to us data relative to personal car accident records so we could combine them with our survey data. Following the construction of a satisfactory path analysis model, a strong prediction (65.6%) of car accident variance was obtained for our young driver cohort. (A)
Samenvatting