Nested-logit mode-choice model for office workers: a case study on Dhaka city

Abstract

This study was an attempt to construct a demand model for transport modes using nested 10git model which can contribute by extending two or mOTechOlce models for the people of Dhaka City. LllVIDEP, transport application software, was used to estimate the mode choice m"del Of1icc workers of Dhaka City were selected for the study, because most of the head offices as well as branch ofllccs of the public and private organizations are located in Dhaka Clly and these offices are a major pull factor for rural-urban migration. The workers of these offices make up a large part of the traffic and are expected to continue their contribution to traffic demand into the future. Data were collected from both secondary and primary sources, About 250 office workers of three private organintions were surveyed randomly by using a pre-tested questiOlUlaire. Office workers chose only five different modes such as pubhc bus, private auto, auto rickshaw, rickshaw and walk as their primary as well as available alternative modes, Soeio economic data included age, gender, educational qlJalifica\lon, status of job, nOll!..hold si~e, and household income. Travel and mode choice data included household car ownership, trip origin, trip destination, frequency of trips, distance from homc to officc in km, most frcquently chosen mode and alternative modes. H also ineluded in-vehicle travel lime, out-vehicle travel time including access time, waiting time and ~gress time together, cost of travcl in Tk., and levcl of scrvice including acecssibility, convenience, privacy, sarety and comfort of primary and di fferent alternative mode8. The outeomes of the generic analysis indicatc that about 52 percent office workers choose public bus as their primary mode. About 86.30% and 81,82% office workers reach their chosen mode on foot and by rickshaw within 10 minutes respectively, Higher income groups choo,oe mainly private auto or auto rickshaw as their primary mode. It is found from the stud)"mat mode choice was also influenced by gender and females were more inclined to choose a comfortable mode,The study tried to indicate a sensible nested logit mode choice modcl for the office workers, but it could not choose any of the model results as a final nested logit model due to wrong sign of different variables and poor statistlcal sib.lifi~ance. One of the most surprising and interesting findings is that there was no rational intcrrelationship between distance, in-vehicle travel time and cost in Dhaka city. Finally, the resnlts obtained in thc specific contcxt led to some recommendations for future users of these pOWerflil modelling tools.