Logistic Regression Analysis of Factors Affecting Travel Mode Choice for Disaster Evacuation

  • Thanawan Phiophuead Kasetsart University
  • Nipawan Kunsuwan Kasetsart University
Keywords: modal split, migration, binary logistic regression, victim, hazard

Abstract

The relationships were analyzed among the factors affecting the travel mode choice between government vehicles and private vehicles used for the evacuation of people in areas experiencing floods and landslides. The relationships were developed using a utility function to predict the probability and proportion for selection of the travel mode in future evacuations based on binary logistic regression. Three models were developed using different analytical factors based on the survey data of a sample group of people in the Mae Pong watershed, Laplae district, Uttaradit province, Thailand. It was found that the factors affecting the selection of travel mode in all three models consisted of sex, household size, families with young members, education, car ownership, experienced a disaster, recognition of shelter location, safety of evacuees while evacuating, reaching the destination quickly, convenience of vehicle access, proportional family management for evacuation, ease of the evacuation procedures of mode, and difference between travel time and walking time to the assembly point. Models 1, 2, and 3 could predict with accuracies of 78.40, 73.46, and 75.30 percent, respectively.

Downloads

Download data is not yet available.

Author Biographies

Thanawan Phiophuead

Department of Civil Engineering, Faculty of Engineering at Kamphaeng Saen, Kasetsart University, Nakhon Pathom 73140, Thailand

Nipawan Kunsuwan

Department of Civil Engineering, Faculty of Engineering at Kamphaeng Saen, Kasetsart University, Nakhon Pathom 73140, Thailand

Published In
Vol 23 No 6, Nov 30, 2019
How to Cite
[1]
T. Phiophuead and N. Kunsuwan, “Logistic Regression Analysis of Factors Affecting Travel Mode Choice for Disaster Evacuation”, Eng. J., vol. 23, no. 6, pp. 399-417, Nov. 2019.