Individuals' spatial social network choice: model-based analysis of leisure-contact selection

Leisure travel holds an important share of the overall amount of travel. However, efforts in transport planning to model and explain leisure travel have been rather limited for a long time. Only recently, a subcommunity of researchers began to use the methods of social network analysis. Existing research in this area, however, has focused on social interaction behaviour of individuals given their social networks. The selection of friends underlying the generation of social networks has received less attention. To fill this gap, our aims in this study are a descriptive analysis of leisure network data, and specification and estimation of a utility-based friendship-decision model. Data on connected personal leisure networks are obtained by using the personal network approach in an extended survey framework. Results offer insights into the decision of who is in contact with whom and which combinations of attributes have either an increasing or decreasing effect on the utility of a leisure relationship. The model can be used to simulate social network generation in an agent-based travel-demand simulation and with that enable the consideration of relevant behavioural patterns in the simulation of leisure trips.