International Choice Modelling Conference, International Choice Modelling Conference 2015

Font Size: 
Social Networks and Trip Behavior in Insular Areas- A Latent Class Model Application
IOANNA KOUROUNIOTI, Amalia Polydoropoulou, Athena Tsirimpa

Last modified: 18 May 2015




The emergence of Social Networks (SN) has modified the way individuals interact with each other and the world and has affected the perception of social relationships such as friendship, information sharing and leisure activities (Kamargianni and Polydoropoulou (a), 2014). In recent years an increasing body of researchers has attempted to explore how SN influence individuals’ personalities and psychologies. However, there is a limited literature on how SN usage affects daily travel behavior. In addition, the relationship between ICT and travel patterns has received a substantial amount of attention, but without focusing on leisure or social travel although it is considered as fastest-growing segment of travel (Van de Berg et al., 2011; Mokhatarian et al., 2006). It is highly probable that the effect of ICT on social travel differs from its effect on travel for other purposes, such as work or shopping.

The aim of this paper is to investigate and quantify the influence of various social networking (SN) usage styles on individuals’ travel behavior, focusing primarily on social/leisure travel. For this purpose a latent class model is under development that consists of two parts: 1. The class membership model, which links the latent SN usage styles to socio-demographic variables; and 2. the class-specific choice model, which is a Poisson regression and show the number of trips made for social purposes of each SN usage style and socio-economic variables.

Latent Class Models (LCM) have been used in many transportation studies in order to quantify transportation behavior (Kamargianni and Polydoropoulou (a), 2014; Ettema, 2010; Walker and Li, 2007; Tawfik and Rakha; 2013; Anowar et al., 2013). According to LCM theory the population can be segmented into a finite number of groups, or classes, that their members share common characteristics and are dissimilar from those in other groups.

The methodology is tested with data from a household activity survey conducted in the island of Chios, Greece in 2014, within the context of GreTIA research project (Green Transportation in Island Areas). Chios is the fifth largest Greek island with a relatively high quality of life, as it is the fourth Greek county in terms of savings (€16,570) and has the third highest car ownership per capita in the country (Hellenic Statistical Authority, 2011). The sample includes 400 individuals that completed two daily activity dairies and provided information on their socio-economic characteristics, the level of SN usage, their gadget ownership and their Internet connection. After the identification of the different SN classes, the number of social trips that the members of each class conduct will be identified.

The innovation of this research covers several topics. First of all, although the effect of ICT on travel behavior has been widely studied the last decade, there are only few surveys that investigate the effect of SN on travel behavior. Secondly, it is interesting to explore the SN usage of remote insular areas inhabitants, since their activity patterns are quite different from urban regions. Furthermore, the findings of this study are expected to assist transport policy makers when making decisions for rural and/or island areas with similar characteristics.




Anowar, S., S. Yasmin, N. Eluru, and L.F. Miranda-Moreno. Analyzing Car Ownership in Two Quebec Metropolitan Regions: Comparison of Latent Ordered and Unordered Response Models. Presented at 92nd Annual Meeting of the Transportation Research Board, Washington D.C., 2013.

Ettema, D. The Impact of Telecommuting on Residential Relocation and Residential Preferences. A Latent Class Modeling Approach. Journal of Transport and Land Use, Vol. 3(1), 2010, pp. 7-24.

Kamargiannai, M. and A., Polydoropoulou, (2014) (a). Generation’s Y Travel Behavior and Perceptions Towards Walkability Constraints among Three Distinct Geographical Areas. Paper presented at the 93rd Annual Transportation Research Board, Washington DC, USA.

Kamargianni, M. and A. Polydoropoulou, (2014) (b). Social Networking Effect on Net Generation’s Trip Making Behavior. Findings from a Latent Class Model. Paper presented at the 93rd Annual Transportation Research Board, Washington DC, USA.

Mokhtarian, P.L., I. Salomon, and S.L. Handy. The Impacts of ICT on Leisure Activities and  Travel: A Conceptual Exploration. Transportation, Vol. 33, 2006, pp. 263-289.

Tawfik, A., and H. Rakha. Latent Class Choice Model of Heterogeneous Drivers' Route Choice Behavior Based on Real-World Experiment. Presented at 92nd Annual Meeting of the Transport Research Board, Washington D.C., 2013.

 Van de Berg, P., T. Arentze, and H. Timmermans. A Latent Class Accelerated Hazard Model of Social Activity Duration. Presented at 90th Annual Meeting of the Transportation Research Board, Washington D.C., 2011.


Walker, J., and J. Li. Latent Lifestyle Preferences and Household Location Decisions. Journal of Geographical Systems, Vol. 9(1), 2007, pp. 77-101.

Conference registration is required in order to view papers.