The effects of social networks on choice set dynamics: results of numerical simulations

Abstract

Now that comprehensive, operational activity-based models of transport demand have become available and are moving to practice, the academic research community has started to address a new challenge: how to develop dynamic activity-based models of transport demand. In this paper, we address one type of dynamic: the formation and dissolution of choice sets. Current activity-based models typically assume that choice sets remain invariant over time, but obviously this is a rather restrictive assumption. In fact, choice sets may change over time because travellers become familiar with new alternatives or may no longer be interested in alternatives they know because they no longer meet the travellers’ expectations or because they have become dominated by other alternatives in the choice set. Knowledge about new alternatives that travellers may decide to explore may be the result of different sources. By implementing activity-travel patterns, travellers may observe alternatives that they have not tried before. Dynamics in choice sets then may come about as the outcome of exploring such new alternatives, resulting from direct visual encounters emanating from activity-travel behaviour. Traveller may also try new choice alternatives due to exposure to advertising, depending on their risk attitude and how well existing alternatives satisfy their needs. Finally, they may decide to explore new alternatives because members of their social network told them about their experiences. The latter is the topic of this paper. The paper will first develop a conceptual framework on the potential influence of social networks on dynamic choice set composition. It is assumed that properties of dyad relationships within social networks will influence dynamics of expectations and preferences. Position in the social networks and the strength of the tendency to belong to the network and comply with its set of attitudes and behavioural patterns will influence the tendency of exploring any new choice options. Based then on experiences, social pressure and possibly other factors, the newly searched and explored alternatives may become part of a travellers choice set, creating the dynamics of interest. Second, based on this conceptual framework, the validity of an agent-based dynamic model of transport demand will be tested and reported.