- Main
A Theory-Based Approach to Travel Behavior Analysis Using Pattern Recognition and Latent Group Identification Techniques
- McBride, Elizabeth Callahan
- Advisor(s): Goulias, Konstadinos G
Abstract
Human behavior is complex: the decisions people make are affected not only by their inner worlds, but also their personal and environmental conditions. In travel behavior research, there is very often a heavy focus on the conditions surrounding behaviors with a lack of focus on the mental processes underpinning the decision-making process. This dissertation presents a novel conceptual model of travel behavior that accounts for these important mental processes and develops data analysis techniques that can help answer research questions based on this framework. In this dissertation, first, the conceptual model is described. Next, an indicator for an important travel behavior concept called fragmentation is established. Common patterns of activities and travel are identified from detailed diaries of travel using sequence analysis, agglomerative nesting, and optimal matching techniques. The relationship between gender and fragmentation in different-sex couples is also analyzed. Then, the design, distribution, and analysis of a survey is presented. This survey looks at travel behavior when COVID-19 stay-at-home orders were in full effect. To the extent possible, the survey used the conceptual model as a guide for what questions to include. A method using latent class analysis to classify travel behavior is also established. Travel behavior attitude are also classified using latent profile analysis. To assess the relationship between attitudes and behavior, the results of these two classifications are compared using cross-classification. Then, a model is created with the goal of following the conceptual model as closely as possible with the data available in the 2017 and 2019 Puget Sound Regional Council Transportation Study. The previously discussed travel behavior classification methods, index of fragmentation, and activity-travel patterns are all used in this model, and it expands upon the research into the relationship between fragmentation and gender. Finally, there is a discussion of issues in travel behavior survey methodology, and suggestions for how to improve these methods.
Main Content
Enter the password to open this PDF file:
-
-
-
-
-
-
-
-
-
-
-
-
-
-