Finally, the four clusters were profiled using multiple discriminant analysis. The discriminating variables were: the travel perceived risks, the youth age, the discrepancy age and the demographic characteristics (the gender, the household incomes, the marital status, the educational level and the self assessed health).
Before achieving the discriminant analysis, the initial set of the nine travel perceived risks has been reduced, by performing a factor analysis. A principal component analysis, involving an Oblimin rotation, was conducted on