Errors in variables in multinomial choice modeling: A simulation study applied to a multinomial logit model of travel mode choice

Modeling travel demand is a vital part of transportation planning and management. Level of service (LOS) attributes representing the performance of transportation system and characteristics of travelers including their households are major factors determining the travel demand. Information on actual choice and characteristics of travelers is obtained from a travel survey at an individual level. Since accurate measurement of LOS attributes such as travel time and cost components for different travel modes at an individual level is critical, they are normally obtained from network models. The network-based LOS attributes introduce measurement errors to individual trips thereby causing errors in variables problem in a disaggregate model of travel demand. This paper investigates the possible structure and magnitude of biases introduced to the coefficients of a multinomial logit model of travel mode choice due to random measurement errors in two variables, namely, access/egress time for public transport and walking and cycling distance to work. A model was set up that satisfies the standard assumptions of a multinomial logit model. This model was estimated on a data set from a travel survey on the assumption of correctly measured variables. Subsequently random measurement errors were introduced and the mean values of the parameters from 200 estimations were presented and compared with the original estimates. The key finding in this paper is that errors in variables result in biased parameter estimates of a multinomial logit model and consequently leading to poor policy decisions if the models having biased parameters are applied in policy and planning purposes. In addition, the paper discusses some potential remedial measures and identifies research topics that deserve a detailed investigation to overcome the problem. The paper therefore significantly contributes to bridge the gap between theory and practice in transport.     

Understanding road network dynamics: Link-based topological patterns

Road network interruptions caused by natural disasters are becoming more frequent and their consequences are becoming of a wider range. The main goal of this work is to identify the most important roads in a network. Herein, a new model is proposed to evaluate the most important roads in the network through the application of biclustering technique, identifying patterns of attributes (road performance measures) and patterns of roads (connectivity patterns). Thereafter the model presented here is compared with the mean geodesic distance variation. Both methodologies are applied to a case study and the pros and cons are discussed as well. Results point out the alpha index as the topological measure more relevant in the normal network flow; moreover the interruption of the links with highest values of connectivity will have larger consequences in the normal functioning of the network than the links with the lowest levels of connectivity. The approach here proposed is a useful insight of the network dynamics, which allows optimizing the worst-case performance of the system. This work can be useful for risk management actors, for civil protection agents, who need to decide on the effective allocation of human and physical resources and define priority areas, and for the government institutions which design the network of facilities.     

A weighted logit freight mode-choice model

The paper describes the estimation of a weighted discrete choice model applicable for analysing choice of mode and crossing for freight in the Oresund region. The study finds that, by applying a weighted logit approach, it is possible to establish a suitable decoupling of agents and shipments. Moreover, by assessing preferences on the basis of a representative baseline OD matrix it is possible to better reflect the dependence between mode substitution effects and geography/infrastructure. The paper presents demand elasticities with respect to monetary cost and travel time as well as value-of-time estimates for five modes and thirteen commodity groups.     

A model for time of day and mode choice using error components logit

The severity of road congestion not only depends on the relation between traffic volumes and network capacity, but also on the distribution of car traffic among different time periods during the day. A new error components logit model for the joint choice of time of day and mode is presented, estimated on stated preference data for car and train travellers in The Netherlands. The results indicate that time of day choice in The Netherlands is sensitive to changes in peak travel time and cost and that policies that increase these peak attributes will lead to peak spreading.     

Detecting urban road network accessibility problems using taxi GPS data

Urban population growth and economic development have led to the creation of new communities, jobs and services at places where the existing road network might not cover or efficiently handle traffic. This generates isolated pockets of areas which are difficult to reach through the transport system. To address this accessibility problem, we have developed a novel approach to systematically examine the current urban land use and road network conditions as well as to identify poorly connected regions, using GPS data collected from taxis. This method is composed of four major steps. First, city-wide passenger travel demand patterns and travel times are modeled based on GPS trajectories. Upon this model, high density residential regions are then identified, and measures to assess accessibility of each of these places are developed. Next, the regions with the lowest level of accessibility among all the residential areas are detected, and finally the detected regions are further examined and specific transport situations are analyzed.By applying the proposed method to the Chinese city of Harbin, we have identified 20 regions that have the lowest level of accessibility by car among all the identified residential areas. A serious reachability problem to petrol stations has also been discovered, in which drivers from 92.6% of the residential areas have to travel longer than 30 min to refill their cars. Furthermore, the comparison against a baseline model reveals the capacity of the derived measures in accounting for the actual travel routes under divergent traffic conditions. The experimental results demonstrate the potential and effectiveness of the proposed method in detecting car-based accessibility problems, contributing towards the development of urban road networks into a system that has better reachability and more reduced inequity.     

Segmenting resident perceptions towards tourism — a cluster analysis with a multinomial logit model of a mountain community

This study investigates aspects of host perception for the formulation of strategic tourism policies. A community segmentation framework with cluster analysis and multinomial logit model was applied to data of a mountain community in northern Italy to reveal the heterogeneity of residents' perception. The results of this study reflect the overall opinions of the sample population and identify homogeneous opinion groups. Positive and negative perceptions about tourism in the region generate a division of residents in environmental supporters, development supporters, protectionist and ambivalent. This study identifies the effect of demographic variation on the clusters and on residents' attitudes towards tourism. Copyright © 2010 John Wiley & Sons, Ltd.     

Network structure and travel patterns: explaining the geographical disparities of road network vulnerability

Inevitably, links in the road network are sometimes disrupted because of adverse weather, technical failures or major accidents. Link closures may have different economic and societal consequences depending on in which regions they occur (regional importance), and users may be affected differently depending on where they travel (regional exposure). In this paper we investigate in what way these geographical disparities depend on the road network structure and travel patterns. We propose aggregate supply-side (link redundancy, network scale, road density, population density) and demand-side (user travel time, traffic load) indicators and combine them in statistical regression models. Using the Swedish road network as a case study, we find that regional importance is largely determined by the network structure and the average traffic load in the region, whereas regional exposure is largely determined by the network structure and the average user travel time. Our findings show that the long-term vulnerability disparities stem from fundamental properties of the transport system and the population densities. Quantitatively, they show how vulnerability depends on different variables, which is of interest for robust network design.     

Decentralizing Pareto-efficient network flow/speed patterns with hybrid schemes of speed limit and road pricing

This paper examines the potential of hybrid schemes of speed limit and road pricing for decentralizing Pareto-efficient flow/speed patterns that minimize total travel time and total emissions simultaneously. Both link flows and speeds are treated as independent variables in our bi-objective formulation. The resulting Pareto frontier is thus weakly dominant to that in previous literature. For any such favorable Pareto-efficient flow and speed pattern, we establish the existence of hybrid schemes of speed limits and non-negative/revenue-neutral tolls, whose set of user equilibrium (UE) solutions contains the Pareto-efficient one, and provide sufficient conditions under which the Pareto-efficient solution can be certainly obtained.     

Advance purchase behaviors of air passengers: A continuous logit model

Modeling the advance purchase behaviors of air passengers is essential when airlines develop revenue management strategies. Therefore, this study empirically investigates advance purchase behaviors based on the air ticket transaction data by using a continuous logit model. The estimation results show that advance purchase behaviors are significantly affected by price, price uncertainty, time of day (morning, afternoon and evening flight), days of week (flight on Friday), months of year (peak or off-peak seasons), and consecutive holiday. Accordingly, different pricing strategies should be used for different flights to maximize revenue.     

How uncertain is the future of electric vehicle market: Results from Monte Carlo simulations using a nested logit model

Plug-in electric vehicles (PEVs) are widely regarded as an important component of the technology portfolio designed to accomplish policy goals in sustainability and energy security. However, the market acceptance of PEVs in the future remains largely uncertain from today's perspective. By integrating a consumer choice model based on nested multinomial logit and Monte Carlo simulation, this study analyzes the uncertainty of PEV market penetration using Monte Carlo simulation. Results suggest that the future market for PEVs is highly uncertain and there is a substantial risk of low penetration in the early and midterm market. Top factors contributing to market share variability are price sensitivities, energy cost, range limitation, and charging availability. The results also illustrate the potential effect of public policies in promoting PEVs through investment in battery technology and infrastructure deployment. Continued improvement of battery technologies and deployment of charging infrastructure alone do not necessarily reduce the spread of market share distributions, but may shift distributions toward right, i.e., increase the probability of having great market success.     

Latent class nested logit model for analyzing high-speed rail access mode choice

This paper explores access mode choice behavior, using a survey data collected in Taiwan. The latent class nested logit model is used to capture flexible substitution patterns among alternatives and preference heterogeneity across individuals while simultaneously identifying the number, sizes, and characteristics of market segments. The results indicate that a four-segment latent class nested logit model with individual characteristics in segment membership functions is the most preferred specification. Most high-speed rail travelers were cost-sensitive to access modes, and thus strategies that reduce the access costs can be more effective than reducing the access times.     

Incorporating spatial equity into interurban road network design

Methods for the road network design problem, typically, are based on optimization of the network efficiency measures (e.g. network-wide travel time) under a predefined budget. In these approaches, equity issues are not taken into account and, consequently, most of the road improvements are planned next to large cities. Thus, disparities between large and small cities increase, which does not conform to sustainable development objectives. In this paper, to overcome concerns associated with traditional methods, equity is incorporated into the interurban road network design problem. To this end, accessibility concepts are employed. However, unlike previous studies, instead of maximizing the total accessibility, a new definition is proposed for inaccessibility, and total inaccessibility is minimized throughout the network. Using this new definition not only is more compatible with the equity issue, but also helps to eliminate the nonlinearity of the problem. Average travel time to neighboring opportunities is utilized to propose this definition for inaccessibility, which captures the reality more effectively. With the aim of this definition, equity is incorporated into the road network design problem implicitly. This is another improvement over previous methods, where a new term in the objective function or a new constraint is added to include the equity. The proposed model is formulated as a mixed integer linear programming (MILP) problem, where the objective is to minimize the aggregate inaccessibility over all the population centers in the network. To illustrate the application of the model, the Northwest region of the United States is used as the case study. The respective exact solution of the example is found using a commercial solver (CPLEX). This new solution is also compared with the solutions from the traditional methods.     

A study on travelers' transport mode choice behavior using the mixed logit model: A case study of the Seoul-Jeju route

This study is to analyze passengers' choice of the mode of transportation when air transportation is in competition with high speed rail (HSR). The Seoul-Jeju route analyzed as an empirical case study, in which the construction of an undersea tunnel to connect Seoul and Jeju city by HSR has been considered. The study also included two new variables, ‘safety of transportation’ and ‘availability of duty free shopping’ in addition to traditional transport choice variables such as travel time, travel costs and frequency of service to reflect special characteristics of the market. As data gathering tools, SP techniques and mixed logit model, for analytical methodology, were utilized. The authors found that the goodness of fit of the models was improved with new variables. The models also showed that the characteristics of business passengers and leisure passengers in choosing the mode of transportation were different. Business passengers were apt to choose a safety secured mode of transportation regardless of fare while leisure passengers preferred to use duty free shops more than business passengers.     

Capacity retirement in the dry bulk market: A vessel based logit model

The paper investigates the effect of vessel specific and market variables on the probability of scrapping dry bulk ships. Using a dataset from 2012 to 2015, we find that the probability of scrapping increases with age, but that the relation between vessel size and scrapping probability varies across the different segments. In addition, while the relation between earnings and probability of scrapping ships is negative, bunker prices seem to affect only the scrapping rate of smaller tonnage. Scrapping probability also increases with an increase in interest rates, freight market volatility and scrap steel prices.     

Modeling aggregate air-travel itinerary shares: logit model development at a major US airline

This study reports the results of aggregate air-travel itinerary share models estimated at the city-pair level for all city-pairs in the US. These models determine the factors that influence airline ridership at the itinerary level and support carrier decision-making. The models are estimated using aggregate multinomial logit methodology and use comprehensive data. Independent variables for the models measure various itinerary service characteristics: level-of-service, connection quality, carrier, carrier market presence, fares, aircraft size and type, and time of day. The results are intuitive, and validation tests indicate that the models outperform existing methods. Finally, the impacts of changing various itinerary service attributes on carrier market share are discussed.     

A qualitative model for road investment appraisal

Recent research in transport appraisal has been predominantly performed by using quantitative linear additive methods such as AHP and MAUT. This paper presents a qualitative model for road investment appraisal based on the DEX method. Qualitative modelling and ability to handle inaccurate and/or incomplete data about options make the DEX method particularly well suited for decision problems involving qualitative concepts and a great deal of expert judgement as is the case in the field of transport. Introduced is a novel and practical way of road appraisal based on the road appraisal framework aimed to support transport policy development in Slovenia.     

Nested logit model formation to analyze airport parking behavior based on stated preference survey studies

The rapid growth of private car ownership and utilization has resulted in significant parking-related congestion problems at large-scale international airports. Parking facilities located in the peripheral areas of airports could be more efficiently used to mitigate congestion and reduce delays due to dramatically increasing traffic demands and inferiorly designed parking lot configurations. In order to analyze the travelers' parking choice preferences for the off-site airport parking activities, a stated preference survey study was conducted around Terminal 3 at Beijing Capital International Airport. Two nested logit models were formulated to analyze the correlation between travelers' parking behavior and their travel characteristics. The research results indicate that travelers prefer to choose off-site parking facilities for long-term parking. A free and fast operation of connections should be provided in consideration of the long-distance layouts of parking facilities. A classification analysis of travelers with different preferences was conducted and the potential changes in travelers' parking choices to off-site parking lots were quantified. The research findings are helpful for traffic management agencies to develop practical guidelines for the establishment of new parking facilities with multiple access points at airports.     

Investigating the associations between road network structure and non-motorist accidents

Road networks channel traffic flow and can impact the volume and proximity of walking and bicycling. Therefore, the structure of road networks—the pattern by which roads are connected—can affect the safety of non-motorized road users. To understand the impact of roads’ structural features on pedestrian and bicyclist safety, this study analyzes the associations between road network structure and non-motorist-involved crashes using data from 321 census tracts in Alameda County, California. Average geodesic distance, network betweenness centrality, and an overall clustering coefficient were calculated to quantify the structure of road networks. Three statistical models were developed using the geographically weighted regression (GWR) technique for the three structural factors, in addition to other zonal factors including traffic behavior, land use, transportation facility, and demographic features. The results indicate that longer average geodesic distance, higher network betweenness centrality, and a larger overall clustering coefficient were related to fewer non-motorist-involved accidents. Thus, results suggest that: (1) if a network is more highly centered on major roads, there will be fewer non-motorist-involved crashes; (2) a network with a greater average number of intersections on the shortest path connecting each pair of roads tends to experience fewer crashes involving pedestrians and bicyclists; and (3) the more clustered road networks are into several sub-core networks, the lower the non-motorist crash count. The three structural measurements can reflect the configuration of a network so that it can be used in other network analyses. More information about the types of road network structures that are conducive to non-motorist traffic safety can help to guide the design of new networks and the retrofitting of existing networks. The estimation results of GWR models explain the spatial heterogeneity of correlations between explanatory factors and non-motorist crashes, which can support regional agencies in establishing local safety policies.     

Arc routing problems to restore connectivity of a road network

After a disaster, restoring accessibility in the affected area is critical for response operations. We study two arc routing problems for clearing blocked roads. The first problem minimizes the time to reconnect the road network, while the second maximizes the total benefit gained by reconnecting network components within a time limit. For each problem, we develop a mixed integer programming formulation and two versions of a heuristic algorithm. We conduct computational experiments on Istanbul data and instances adapted from the literature. The heuristics achieve near-optimal or optimal solutions quickly in most of the tested instances.     

Interregional trade network analysis for road freight transport in Greece

This paper suggests a flexible decision support framework for the strategic planning of a freight transport hub network in Greece. The proposed methodology treats practical aspects related to the optimal number, location and geographical covering of hubs, through the network analysis of interregional trade, based on original survey data for road freight flows during 2004–2012. The results offer insights into the hierarchical structure of the network and related investment priorities, as the hub role of a prefecture is found to be strongly influenced by high population densities and manufacturing specialization, and its location along highway corridors.