A large-scale study of possibilities for and social benefits of high speed rail (HSR) in Norway has recently been conducted. Following this, the subject of HSR has been frequently debated in Norwegian media. An important part of the cost-benefit analyses for HSR is the predicted ridership. Discrete choice modeling is the conventional method for estimating the mode choice probabilities used in these forecasts. Historically, the covariates taken into account in such models are attribute values for each modal choice as well as socio-economic attribute values for the travelers. However, even conditional on these variables there is often a high degree of individual, unobserved heterogeneity which contributes to low explanatory power. This is a potential problem, especially in the context of forecasting. This thesis tries to capture some of the individual heterogeneity by including personality traits as latent variables in the utility functions in the choice models. These personality traits are mainly revealed through indicator variables in the form of questions regarding attitude and behaviors in everyday life. This can for instance be information regarding recycling behavior to reflect environmental consciousness, or information regarding safety behavior in traffic to reflect the preference for safety. The obvious advantage of such indicators is that information not inferable from market behavior can be included in the decision making process.
The thesis consists of two parts. The first part is a complete analysis of the covariance structure of the indicators I have available by means of factor analyses. Based on this I provide suggestions for how personality traits should be designed and also establish the link between these personality traits and observable characteristics as income, gender and age. The second part is integrated latent variable and choice models, where the personality traits ``comfort'' and ``global environmental consciousness'' are included as latent variables to explain the choice between air transport and HSR in Norway. The market segment on which I focus is business travels on the links Oslo-Bergen and Oslo-Trondheim and the analysis is based on a stated preference study. I find that both these personality traits are significant. Moreover, they affect the choice probability for HSR positively and seem to do a better job in explaining mode choice than the available observable individual specific characteristics. I am cautious when drawing conclusions from the models since they are simple in terms of specification of utility functions. However, they shed light on aspects important for the utility of HSR that are easily forgotten in conventional analyses. This includes in particular the heterogeneity in how individuals' utilities are affected by changes in comfort, and the ``purchase of moral satisfaction'' by traveling more environmentally friendly.
Finally, an important contribution of this thesis is that it summarizes the state of the art theories related to such analyses. It is to my knowledge no other sources in which theories regarding factor analyses, discrete choice models, latent variable models and a consistent framework in which latent variables enter the choice model are collected. In this manner my thesis provides added value for researches wanting to analyze choices in an attitudinal context since it describes the complete theoretical foundation of all the related processes.