Development of Hybrid Models of Teenager’s Travel Behavior to School and to After-School Activities

Substantial changes in lifestyles, urban environments and transportation systems have led to changed physical activity patterns, especially among underage people. Although we may know the demographic and economic characteristics of underage students’ families and the communities where they live and attend school, we have little scientific evidence of the individual teenagers’ (12 to 18 years old) activities, travel behavior and attitudes. In traditional societies there was comparatively little discrepancy between adolescents and adults because they grew up in comparable worlds. However, with rapid change and social media bringing the outside world into teenagers’ lives, larger generational differences are emerging. This thesis contributes to the understanding of various factors that affect teenagers’ travel behavior. As a first step, we analyze teenagers’ activity patterns and time use in school days and in Saturday and the transport mode that they use, in order to identify their travel needs. The results indicate that teenagers conduct a number of trips, especially after-school trips, without the supervision of their parents, while the mode use patterns significantly differ among the trip purposes and among distinct geographical areas. As current teenagers spend significant amount of time on online social networking (OSN), we further analyze how much, why, and how teenagers utilize social media, and how its usage affects their travel behavior. Latent Class Poisson Regression models are developed in order to identify teenagers’ trip making behavior for social purposes of various OSN usage styles, while the results indicate that those who use OSN in a rational or addictive way, conduct more social trips than those who are indifferent to OSN, thus OSN does not substitute face-to-face communication. The developed framework offers significant insights to researchers for the data required in order to model the relationship between OSN and trip making behavior. The thesis is also concerned with investigating the effect of social influence on decision making and more specific the effect of parents’ walking patterns on teenagers’ attitudes towards walking and mode choice behavior. We present a methodological framework that incorporates social interaction effect into Hybrid Choice Models (HCM) and provide the required mathematical equations. The model estimation results indicate that, if the teenagers perceive that their parents are walking-lovers, then this increases their probability of loving walking too. Even though the application focus on teenagers, the framework is general and can be applied to modeling adults’ behavior as well.