In the context of route choice, modeling the process that generates the set of available alternatives in the mind of the individual is a complex and not fully explored issue. Route choice behavior is influenced by variables that are observable, such as travel time and cost, and unobservable, such as attitudes, perceptions, spatial abilities, and network knowledge. In this study, attitudinal data were collected with a web-based survey addressed to individuals who habitually drive from home to work. The paper pro- poses a methodology to conduct a proper application of factor analysis to the route choice context and describes the preparation of an appro- priate data set through measures of internal consistency and sampling adequacy. The paper shows that, for the data set obtained from the web-based survey, six latent constructs affecting driver behavior were extracted and scores of each driver on each factor were calculated. An insightful representation of individual route choice behavior should answer the following two questions: • Which routes does a person consider as travel alternatives? • Which route does a person choose among them? Most route choice models focus on the answer to the second ques- tion by calculating the probability of choosing a route over alternatives selected with heuristic techniques or simulation methods. Choice set generation models concentrate on the answer to the first question, but modeling the process that generates the set of available alternatives in the mind of the individual is a complex and not fully explored issue. According to Ortuzar and Willumsen (1), this topic is the most dif- ficult to resolve in the context of modeling revealed choice behavior. Stopher (2) and Williams and Ortuzar (3) showed that a statistical inconsistency of the utility parameter estimates is the result of a biased construction of the choice set. Swait and Ben-Akiva (4, 5) found in model parameters differences related to the composition of the choice set. Swait (6 ) recently developed a choice set generation model belonging to the generalized extreme value family. These studies concentrated on mode choice, where the number of possible alternatives is relatively small and the issue is mainly a mat- ter of perception. This research focuses on automotive route choice, where individuals face a large number of routes and the issue is also a problem of awareness. From the analyst’s perspective, travel time and cost are the vari- ables commonly considered for modeling individual route choice behavior. Relevant contributions from fields such as psychology and geography provided insight into the human decision-making process and gave evidence that unobservable differences among travelers influence their travel patterns, suggesting that an enhanced repre- sentation of the choice set generation process should include un- observable factors such as attitudes, perceptions, spatial abilities, and network knowledge. This paper introduces an exploratory analysis of latent factors influencing choice set formation in individuals who habitually drive from home to work. Answers to questions related to drivers’ atti- tudes, spatial abilities, and driving preferences, as well as stated choices of the routes considered for traveling from home to work, were collected with a web-based questionnaire submitted to faculty and staff members of Turin Polytechnic, Turin, Italy. The estimation of a choice set generation model, relating the stated choice sets to indicators of the latent factors extracted, is a possible further devel- opment of this research. In this direction, Ben Akiva and Boccarà (7 ) presented a framework for discrete choice models with latent choice sets and estimated a probabilistic choice set generation model for mode choice. This paper focuses on the methodology of gathering attitudinal and behavioral data through a web-based survey, the measure of internal consistency and adequacy to factor analysis of the data set collected, and the extraction of latent factors and indicators suitable to the estimation of a choice set generation model. This paper is structured as follows. The next section introduces the theoretical background of the research. The third section illus- trates the survey design and data set handling. The fourth section presents significant results of statistical and factorial analysis. The last section presents the conclusions and addresses further efforts in this research. THEORETICAL BACKGROUND A review of the literature in route choice and choice set generation modeling, which includes contributions by psychologists and geog- raphers about spatial behavior and way-finding techniques, de- fines the context of this research. A review of statistical methods, Methodology for Exploratory Analysis of Latent Factors Influencing Drivers’ Behavior Carlo Giacomo Prato, Shlomo Bekhor, and Cristina Pronello C. G. Prato, and C. Pronello, Turin Polytechnic, Department of Hydraulics, Transport and Civil Infrastructures, Corso Duca degli Abruzzi, 24, 10129 Torino, Italy. Current affiliation for C. G. Prato: Transportation Research Institute, Technion–Israel Institute of Technology, Technion City, 32000 Haifa, Israel. S. Bekhor, Technion–Israel Institute of Technology, Faculty of Civil and Environmental Engineering, Haifa 32000, Israel. 115 Transportation Research Record: Journal of the Transportation Research Board, No. 1926, Transportation Research Board of the National Academies, Washington, D.C., 2005, pp. 115–125.