Immigration, residential location, car ownership, and commuting behavior: a multivariate latent class analysis from California |
| |
Authors: | Jarad D Beckman Konstadinos G Goulias |
| |
Institution: | (1) Department of Geography, University of California Santa Barbara, 1832 Ellison Hall, Santa Barbara, CA 93106, USA |
| |
Abstract: | Using latent class cluster analysis, this paper investigates the spatial, social, demographic, and economic determinants of
immigrants’ joint distribution among travel time, mode choice, and departure time for work using the 2000 Census long form
data. Through a latent tree structure analysis, age, residential location, immigration stage, gender, personal income, and
race are found to be the primary determinants in the workplace commute decision-making process. By defining several relatively
homogeneous population segments, the likelihood of falling into each segment is found to differ across age groups and geography,
with different indicators affecting each group differentially. This analysis complements past studies that used regression
models to investigate socio-demographic indicators and their impact on travel behavior in two distinct ways: (a) analysis
is done by considering travel time, mode choice, and departure time for work simultaneously, and (b) heterogeneity in behavior
is accounted for using methods that identify different groups of behavior and then their determinants. Conclusively the method
here is richer than many other methods used to study the ethnically diverse population of California and shows the addition
of geographic location and latent segment identification to greatly improve our understanding of specific behaviors. It also
provides evidence that immigrants are as diverse as the non-immigrant population and transportation policies need to be defined
accordingly.
|
| |
Keywords: | Travel behavior Immigrants Latent class cluster analysis Commuting |
本文献已被 SpringerLink 等数据库收录! |
|