MODELING COMMUTER PREFERENCES FOR A BUS

European Scientific Journal July 2015 edition vol.11, No.21 ISSN: 1857 – 7881 (Print) e - ISSN 1857- 7431 MODELING COMMUTER PREFERENCES FOR A BUS Eric Nyarko, M. Phil Caleb Owusu-Ansah, M. Phil Charles Kwofie, M. Phil Department of Statistics, University of Ghana, Legon, Ghana Abstract The research deals with the use of the discrete choice experiment technique and Random Utility Theory (RUT) to analyze bus choice behaviour for commuters in Asafo-market in Kumasi, Ghana. The principal aim of the study was the calibration of a logit model to forecast consideration sets. In order to estimate the model parameters, a specific survey was carried out inside the urban area of Kumasi. The survey focused on passengers and involved mainly employees, self employed workers and students (120 respondents). The data collected from the experimental survey was further calibrated and segmented according to gender. The magnitude of estimates generally indicates that commuters highly value buses with fixed departure time, spacious seats and Air Condition (AC). However, an increase in fare level will result in a disutility of bus choice. Keywords: Bus, discrete choice experiment, discrete choice modeling, Random Utility Theory Introduction Transport is an indispensable element of development and socioeconomic growth. As engine of economic integration, transport infrastructure and services facilities constitute a precondition for facilitating trade and the movement of goods and person. Long perceived as a tool for accessing national and regional trade in a radically changing global environment, transport infrastructure remains a pillar of development with a view to accelerating growth and reducing poverty (Okoko, 2006). Road transport is the predominant means of travelling in Ghana, which enhances high passenger travels and carting of goods and services. It provides essential role by linking the country to others in the entire West African sub-region. Transportation has developed rapidly in Ghanaian societies, but there is competition between privately owned cars and commercial vehicles. Despite 420 European Scientific Journal July 2015 edition vol.11, No.21 ISSN: 1857 – 7881 (Print) e - ISSN 1857- 7431 the high growth rate in urban centers like Accra and Kumasi, there have been some shortfalls in public policy. This has contributed to longer shuttling period and journey delays, lengthy waiting times for commercial vehicles both at and between terminals, high accident rates, and localized poor air quality (Afful, 2011). However, as a result of the poor quality of travel in bus transportation systems in Ghana with a declining trend in commuters’ choice of buses, policy-makers and transport operators are constantly in search of solutions for improving bus choice, especially in urban areas. In fact an increase in Bus use, with a concurrent reduction in the use of private cars, could help to reduce many problems like traffic congestion, air and noise pollution, and energy consumption. For these reasons, several works have been made by various studies on urban public transport; for example, Baidoo, Nyarko and Mettle (2015) in their study of modeling mode choice in passenger transport with discrete choice experiment revealed that in choosing a commercial vehicle, commuters generally took into consideration their safety, travel distance comfort and waiting time before making their choices. Travel safety is highly valued by commuters. Pavlyuk and Gromule (2010) in their study considered three transport options; car, coach, and train. A nested discrete choice model was used to analyze factors that influence passengers’ choice. The authors concluded that departure time had a significant influence on bus/train choice. Passengers who choose price as a key factor in their selection prefer to use the train. The terminal point as a destination predictably increases the probability of train selection. Baidoo and Nyarko (2015) examined mode choice between bus and private car, with the habit of using a bus being one of the attributes. Binary logit model and its marginal effects were employed to assess commuters’ behaviour with regards to their choice between different transportation modes in traveling to Accra central. The authors concluded that the level of noise, comfort, and time (morning trips) will result in a disutility of public transport choice. Van der Waerden, Borgers, Timmermans, and Berenos (2007) used MNL models to examine the choice between car, bus and bicycle for different journey purposes. They argued that the cost and time attributes dominate, obtaining a seat is significant across journey purposes. Baidoo and Nyarko (2015) employed Discrete Choice Experiment (DCE) and Random Utility Theory (RUT) to measure service quality in public transport. Probit model was calibrated and segmented based on gender. They concluded that an increase in the walking distance to bus stop and transport fare will result in a disutility of service quality attributes. Catalano, Lo Casto and Migliore (2008) employed random utility model to analyze travel mode choice behaviour for commuting urban trips in Palermo, Italy. The authors found out that, for the specific case of Palermo, the multinomial logit proved to be the best urban 421 European Scientific Journal July 2015 edition vol.11, No.21 ISSN: 1857 – 7881 (Print) e - ISSN 1857- 7431 transport demand model, even if the choice set contained three car alternatives. However, as far as the authors are concern, most of the studies that make use of DCEs are carried out in the Western world with paucity of information on users’ attitude when they have a mode of choice between buses. In this study, a Discrete Choice Experiment (DCE) and Discrete Choice Modeling (DCM) which is rooted in Random Utility Theory (RUT) are used to estimate bus passengers’ attitude towards endogenous consideration sets. This will help to propose policy intervention issues in urban areas in developing countries. Methods and Materials Sample and Data Collection Procedure The data were collected using Computer Assisted Personal Interview (CAPI). This explains a large response rate for a Discrete Choice Experiment (DCE). Commuters were sampled using simple random sampling techniques for the study. Commuters who board buses from the Kumasi-Asafo bus station to various destinations (Cape Coast and Accra, for example) and owned or have access to a private car were targeted since the study sought to analyze the hypothetical choice of a bus by these people. A sample size of 120 respondents was chosen for the study. A total sample of fifty (50) individuals each with 16 choice sets and fully generic parameter specification for design attributes and covariate effects might just be acceptable for choice experiment (Hensher et al., 2005). Revealed/Stated Choice Design The basic shortcomings of SP surveys are not present in RP surveys as they deal with existing actual situations being experienced (...truncated)