ANALYSIS OF TRANSPORTATION MODE CHOICE OF THE PASSENGERS USING MULTINOMIAL LOGIT MODEL FOR WORK AND SCHOOL TRIPS: A CASE STUDY IN HAWASSA CITY

Abstract

The urbanization and urban sprawl of the city are becoming expanded in Ethiopia. People migrated to urban centers like Hawassa city in order to work, trade, and learn. Due to this, the transportation planners and engineers are always encountering a problem on how to provide adequate transportation services and facilities within the city because of the various mode choice behavior of the travelers. Hence, this study focused on identifying factors affecting mode choice behavior, analyzing the work and school trip makers’ mode choice behavior in Hawassa city using the Multinomial Logit Model by determining and forecasting the current and future mode share respectively. The transport modeling process as a travel demand forecasting tool used in the analysis of probable market shares of different motorized modes. The model developed with questionnaires distributed for this purpose by using workplace and school survey in the seven sub-cities of the Hawassa city. The primary data, like total waiting time were collected through site investigation at different major station of the city. Likewise, the secondary data like travel cost within each station gathered from the Hawassa City Road and Transportation Bureau. SPSS Software was used to process data for a Multinomial Logit Model (MNL) and to explain the travel patterns and mode choice of employees and students. The model was made with mode used in a trip as dependent variable and out of the vehicle travel time, in vehicle travel time, travel cost, income and comfort as an independent variables. Results indicated that the factors that significantly affect the choice of transport modes were average monthly income, in-vehicle travel time, out of the vehicle travel time, total travel cost and comfort during journey. The prediction level of the developed model was 79.5% for the employees and 82.9% for the students. The employee’s current probabilities of the modal split for city bus, damas, Bajaj and motorcycle were 28.83%, 25.4%, 6.2% and 39.58% while for the students were 35.92%, 29.74%, 19.79% and 14.53% respectively. The validation and verification of the developed model was done by testing the reasonableness of the sign of the parameters and by examining it with the three fundamental properties of the MLM. The estimations of the model might also be valuable to policy-makers and transport planners, who can estimate possibilities for various modes and determine the impact of unique policy modifications on the need for trip makers