Social networks' group tie strength and travel behavior

Abstract The last decades saw an increased interest in the social aspects of urban mobility, particularly in understanding the linkage between social network (SN) ties and travel behavior. Most studies analyzed ego-centric networks, focused on dyadic relations and distinguished between weak and strong ties. The purpose of this study is to understand the hidden layers connecting social network group ties and its associated travels within a meeting context. Data included 315 relations based on 137 group meetings that took place in the city of Tel Aviv, Israel at either a cafe, pub or a restaurant and involved three or more participants. We developed a group tie strength index as a spectrum ranged from weak to strong ties based on four parameters: intimacy, duration of connection, frequency, and modes of communication among group members. The study reveals that majority of the connections tend to be at the medium of the spectrum. This indicates that the common classification into weak/strong ties in relation to travel behavior is overlooking important segments of ties and that there is a difference in tie strength between dyadic and triadic/group meetings. Classification and Regression Trees (CART) model analysis explored how Social Network group tie strength (SNGTS) of members participating in a meeting event can be predicted by the event context and the contextual travel pattern. Two CART trees were found. Each started with a different variable for the first split differentiation and showed different tree paths for explaining the group tie strength for an event-based participants. These include the “communication means” that were used to arrange the meeting and the “meeting type” (i.e.- friends, family, and business- oriented meetings) variables. Five more variables were found to be related to predicting the SN group tie strength including: meeting duration, distance of meeting place from residence, transport means and travel duration to the meeting and age.