While young motorcyclists are more likely to conduct risky driving behaviors and be involved in crashes, not all are the same. Past studies have examined such heterogeneity by grouping drivers and then assessing differences between the subgroups solely on variables or constructs such as drivers' personality traits, attitudes or risky behaviors. However, the discrepancy of causal structure relationships between personality traits and risky driving behavior in different subgroups still needs to be explored further. This study thus proposes a multi-group analysis based on structural equation modeling to investigate the structural discrepancy that may exist in distinct groups. On the basis of cluster analysis of personality traits, this work identifies four types of young motorcyclists in Taiwan. Findings show that significant difference on a single construct could result in significantly different causal structures. Sensation seeking, compared to other personality traits, has the highest effect on risky riding behavior forall clusters; nevertheless, sensation seeking causes risky riding behavior following various causal paths in distinct clusters: In the risky cluster, riding confidence and perceived utility mediates between sensation seeking and risky behavior while in the aggressive cluster, sensation seeking directly connects to risky behavior. This research also finds discrepancies in other personality traits between clusters and discusses intervention strategies corresponding to the different features of causal relationshipsin each cluster.
Abstract