Deriving Constructs for Hierarchical Information Integration Experiments in Regional Public Transport

With the (Integrated) Hierarchical Information Integration approach a larger number of attributes can be included in choice or rating experiments since similar attributes are summarised by higher order constructs. The aim of the paper is to derive such constructs in the context of service quality in regional transport and to determine indicators of validity. To that end sorting data is collected in interviews with train users and analysed at aggregate level using Multidimensional Scaling, leading to five clusters. Rand indices are calculated to compare the aggregate solution with data at individual level. Results show that the sorting performed by both train and car users is rather well represented by the aggregate solution.