01015548

Component

https://doi.org/10.3141/1921-09

http://scholar.google.com...lker&publication_year=2005

Discrete choice analysis has become an industry standard in land use and transportation models. Such models are fundamentally grounded in individual choice; therefore, the treatment of interdependencies among decision makers is a formidable challenge. Through an empirical application to mode choice, the capture of interdependencies in discrete choice is described and illustrated. Decision makers are assumed to be influenced, for example, by people of similar socioeconomic status who are nearby. Given such social and spatial network relationships, the choice model captures interdependencies in two ways: (a) including in the systematic utility variables that describe choices of others in the decision maker’s social and spatial network and (b) allowing for correlation across the disturbances of decision makers within the same social and spatial network. Variations of these approaches (including their combination and the use of random parameters) are tested with mode choice and compared with traditional methods of market segmentation. The application results indicate that the proposed methods for capturing interdependencies are significant and superior to traditional methods. Furthermore, capturing the interdependencies in the systematic utility is sufficient: it is better than the model with just correlation and is not significantly worse than the model with both the systematic term and correlation. The systematic term also captures a feedback effect that can propel the adoption of a new mode over time, for example. Models are estimated with the use of a traditional transportation data set and readily available software.

01015546

English

pp 70-78

2005

Transportation Research Record: Journal of the Transportation Research Board

0309093953

Print

Figures (2) ; References (46) ; Tables (1)

Choice models; Decision making; Mode choice; Socioeconomic factors

Discrete choice analysis; Interdependence

Economics; Highways; Planning and Forecasting; Society; I72: Traffic and Transport Planning

TRIS, TRB, ATRI

Jan 18 2006 3:39PM

• Analysis of Joint Household Decisions About Car and Motorbike Ownership and Use in Taiwan
• Development of a Scheduling Model for Car Tourists’ 1-Day Tours
• Efficiently Estimating Nested Logit Models with Choice-Based Samples: Example Applications
• Escorting Children to School: Statistical Analysis and Applied Modeling Approach
• Evaluation of Optimization Methods for Estimating Mixed Logit Models
• Florida Activity Mobility Simulator: Overview and Preliminary Validation Results
• Incorporating Trip-Chaining Behavior into Network Equilibrium Analysis
• Market-Based Framework for Forecasting Parking Cost in Traditional and Microsimulation Modeling Applications
• Path Size Modeling in Multimodal Route Choice Analysis
• Simulation Estimation of Mixed Discrete Choice Models with the Use of Randomized Quasi-Monte Carlo Sequences: A Comparative Study
• Testing Mixed Logit and Probit Models by Simulation
• Tests of a Family of Trip Table Refinements for Long-Range, Quick-Response Travel Forecasting
• Travel Demand Modeling for Regional Visioning and Scenario Analysis
• Urban Commercial Vehicle Movement Model for Calgary, Alberta, Canada