ABSTRACT
This chapter aims to explain the theoretical foundations behind the models based on the random utility theory. Some of their properties are studied to understand how to use them appropriately as sub-models within a wider Land Use–Transport Interaction Models (LUTI). The random utility models assume the most frequently used theoretical paradigm to represent the choices made between the discrete alternatives. Random utility models are not born as models designed to study cases of spatial dependence between alternatives. However, they are widely used in LUTI modelling, mainly because of their capacity to study user behaviour, especially in terms of their location choices. As the study of spatial relationships is one of the basic characteristics of LUTI models, the simplest utility models, such as multinomial logit model, fail when they consider that the errors associated with the utility functions of each choice alternative that are distributed independently and identically.
