Application and adaptation of the CNET interview protocol for the construction of individual decision networks of complex leisure travel decisions

Abstract

As the need for leisure time increases in the developed countries, the number of trips that are generated by this type of activity is also increasing (Kiiskilä & Kalenoja, 2001). Therefore, travel choice behavior of day trippers in congested urban areas should get more attention from decision makers in order to promote low-impact and sustainable transportation modes and hence to reduce the pressure on urban infrastructure. However, a prerequisite of TDM that are implemented by policy makers for being both effective and efficient, is that more research is done in advance about individuals’ choices and their complex decision processes (Den Hartog et al, 2005). In order to contribute to this, the research addresses the application of the CNET (Causal Network Elicitation Technique) interview method developed by Arentze at al (2007) for the assessment of complex leisure travel decisions. Leisure travel decisions may be more elaborate and heterogeneous than regular travel decisions, which further strengthen the need to understand individuals’ Decision Network (DN). This method will be tested in the case of fun shopping activity in Hasselt (Belgium), in order to find individuals’ reasoning behind activity, shopping location and transport mode decisions. The interview method enables the modeling of individuals’ conscious choices and decisions in a DN shaped as a Bayesian Inference Network (BIN). In this research, several different possible methods to construct BIN are examined. These methods are qualitative structured interview versus pick-list for learning the network structure of BIN on the one hand, and scaling, ranking, AHP and SP method (FFD) for learning the parameters of BIN on the other. Results show a significant difference among the DNs generated by different methods in terms of its complexity. The research has revealed the most efficient combination of interview methods to elicit individuals’ mental representation of the decision problem.