Dogit Model Research Articles

Modeling mode choice of customized bus services with loyalty subscription schemes in multi-modal transportation networks

This study models the equilibrium mode choice of emerging customized bus (CB) services in a multi-modal urban transportation network. Two unique characteristics of CB services are considered: (1) the reservation of a seat for every CB passenger, and (2) the loyalty scheme for long-term CB subscribers. The first characteristic eliminates in-vehicle crowding but imposes tight constraints on CB demand, while the second characteristic encourages passenger loyalty to CB services. An advanced equilibrium mode choice model based on the dogit model, a well-established travel choice model that is widely used to consider captive mode choices, is proposed to account for the choice behavior of passengers loyal to CB services. In contrast, a nested weibit model is used to account for the choice behavior of passengers not loyal to CB services by simultaneously considering the effects of mode similarity and mode heterogeneity on mode choice behavior. The proposed model is formulated as an equivalent mathematical programming problem, which guarantees the existence and uniqueness of solutions. Numerical experiments are conducted to illustrate the properties and applicability of the proposed model in a real-world setting. The results demonstrate the importance of considering passenger loyalty and managing mode similarity and heterogeneity when modeling emerging CB services. The results also show the applicability of the proposed model for evaluating the performance of a network with CB services and the design and pricing of CB operation and subscription schemes.

Dogit model and rational inattention

This paper establishes an equivalence result between the Dogit model and the rational inattention model. The Dogit choice model introduced by Gaudry and Dagenais (1979) is an extension of the multinomial logit model which can potentially violate the independence of irrelevant alternatives axiom and allow for captive or inattentive behaviors. McFadden (1981) argues that a drawback of the Dogit model is its inconsistency with the standard random utility framework. We show that when the information cost is modeled using a generalized Shannon entropy function, the Dogit model is observationally equivalent to a rational inattention model.

A multi-modal network equilibrium model with captive mode choice and path size logit route choice

Abstract In this paper, we consider captive mode travelers (those who have no other choices but rely on one specific travel mode for daily commuting trips) in a multi-modal network equilibrium (MMNE) problem. Specifically, the dogit model is adopted to account for captive mode travelers in the modal split problem, and the path-size logit (PSL) model is used to capture route overlapping effects in the traffic assignment problem. The dogit-PSL MMNE model is formulated as an equivalent entropy-based mathematical programming (MP) problem, which admits solution existence and uniqueness. Three numerical examples are provided. The first example examines the effects of mode captivity and route overlapping on network performances and observes that accounting for captive mode travelers would produce different equilibrium states and hence the network performance indicators. The second example applies the dogit-PSL MMNE model for evaluating the exclusive bus lane (EBL) expansion plans, in which a consistent synthetic proportional index is proposed. Numerical results show that considering mode captivity may produce substantial impacts on the odds (up to 50 percent of odds in the given scenarios) of making different EBL line expansion decisions. The third example implements the dogit-PSL MMNE model in the Seoul network to show the applicability of the dogit-PSL MMNE model in a real-size multi-modal system.

Implementation of a new network equilibrium model of travel choices

Abstract This paper develops a new combined network equilibrium model by using more behaviorally sound mathematical forms to represent the four travel choices (i.e., trip frequency, destination, mode, and route) in a conventional travel demand forecasting process. Trip frequency choice relates to the traveler decision on “making a trip” or “not making a trip” so it is given by a binary logit model. Destination choice is formulated as a parameterized dogit model of which the captivity parameters (expressed as functions of independent variables) allow individual travelers to be captive to specific destinations. Mode choice is given by a two-level nested logit model to avoid IIA restriction. Trip assignment is based on Wardrop's “user-optimized” principle. All model forms describing travel choices are in response to the level of services incurred by the transportation system. Through the introduction of inclusive values, the traveler decisions concerning trip frequency, destination, mode, and route choices are inherently interrelated and jointly determined. To obtain solutions of the new combined model, it was reformulated as an equivalent convex programming problem with linear constraints, a great advantage from the computational aspects. The model was applied empirically to a transportation network in New Jersey. The application results show that the new model is consistently better than the commonly used logit combined model in reproducing the observed trip flows from origin zones, origin to destination (O-D) trip flows, O-D trip flows by mode, and trip flows on the network links.

Stated Responses to Policy Interventions and Technological Innovation of Electric Motorcycles in Laos

Recently, the heavy pollution from motorcycles prompts the trend to promote electric motorcycles (EMs) with no emission and higher efficiency. This research represents people’s stated responses to different policy interventions and technological innovations of EMs, from the perspective of human choice behavior, in Vientiane, Laos. Based on a stated preference (SP) survey, a dogit model with parameterized captivity functions was adopted to examine the important influential factors to the choice of EMs by accommodating people’s heterogeneous captivity to different alternatives in a choice set. Simulations under different policy-related and technology-related scenarios were conducted, as well. Based on the analysis results, it is found that the preferred distance to charge station is 0.8 km - 1.0 km and people are more likely to choose EMs when the cruising range is longer, operation cost is lower, and the diffusion rate is lower.

Network Equilibrium Model with Dogit and Nested Logit Structures

This paper develops a combined equilibrium model for the simultaneous prediction of trip distribution, mode choice, and trip assignment. Trip distribution is formulated as a dogit model to take into account travel behavior of compulsory (work) and discretionary (nonwork) trips. Mode choice is given by a two-level nested logit model to avoid the independence from irrelevant alternatives restriction. Trip assignment is based on Wardrop's user-equilibrium principle. The log sums of the mode choice model were computed and fed back into the trip distribution step. Under this iterative process, destination choice of discretionary trips would respond to changing travel cost conditions occurring in mode choice and trip assignment steps, whereas destination choice of compulsory trips would remain fixed in the study time period. The proposed combined trip distribution, mode choice, and trip assignment (CDMA) model was reformulated as an equivalent convex programming problem with linear constraints and applied empirically to a county-level network in New Jersey. The results indicate that the CDMA model is better than the commonly used logit combined model in reproducing the observed link flows, origin–destination (O-D) trip flows, and O-D trip flows by mode.

Analysis of Paratransit Drivers' Stated Job Choice Behavior under Various Policy Interventions Incorporating the Influence of Captivity: A Case Study in Jabodetabek Metropolitan Area, Indonesia

This study analyzes paratransit drivers' stated job choice behavior under various policy interventions. First, the current profiles of four typical types of paratransit drivers (becak, ojek, bajaj, and angkot drivers) are clarified in Jabodetabek Metropolitan Area, Indonesia. Second, factors affecting their future driver job choices under policy interventions from social and environmental aspects were investigated based on a stated preference (SP) survey, conducted in 2010. In the SP survey, future driver jobs are assumed by considering salary level, employment opportunity, employment status, operation cost, subsidy for low-emission vehicles, vehicle fuel types, and so on. Third, direct evaluation of different factors on job choices were conducted by cross-tabulations. Dogit model was further applied to properly represent the captive job decision. The model estimation results confirmed that drivers are relatively captive to ojek and bajaj, and subsidy for low-emission vehicles is influential to drivers' stated job choices about ojek and bajaj.

Public Transportation Share Ratio Prediction Based on Disaggregate Dogit Model

In order to analyze the distributing condition of urban passenger flow scientifically and correctly, the disaggregate Dogit model is presented to predict the public transportation share ration in city, which is carried out by means of analysis of the outer and inner factors that affect the choice of modes of transportation and is based on the random utility theory. Establishment of the model, parameter identification and the process of calculation are described in detail. Finally, according to the proposed algorithm, the public transportation share ratio forecast test is carried out using the field survey data. The results of independent sample test indicate that the model has a finer precision and stability.

Distribution and Assignment of Compulsory and Discretionary Traffic

This paper presents the development of a combined equilibrium model for the simultaneous prediction of the destination and route choices that face suburban or rural automobile travelers. Destination choice is given by a dogit model to take into account the travel behavior of compulsory (work) and discretionary (nonwork) trips. A logit-based route-choice model was employed, with a stochastic user equilibrium principle, to develop route flows between each origin–destination pair. The natural logarithm of the denominator of the logit route-choice model (i.e., log-sum) was computed and fed back into the destination-choice step. Under this iterative process, the destination choice for discretionary trips responded to changing travel-cost conditions, whereas the destination choice for compulsory trips remained fixed in the study time period. The proposed combined destination- and route-choice (CDR) model can itself be reformulated as an equivalent convex programming problem with linear constraints, a great advantage from a computational perspective. The CDR model was applied empirically to a county-level network in New Jersey. The results encourage further applications of the CDR model to large-scale networks.

MODELING THE INCIDENCE OF SELF-EMPLOYMENT: INDIVIDUAL AND EMPLOYMENT TYPE HETEROGENEITY

Modelling the incidence of self-employment has traditionally proved problematic. Whilst the individual supply side characteristics of the self-employed are well documented, we argue that the literature has largely neglected demand-side aspects. We explore the determinants of self-employment using individual level data drawn from the U.S. Survey of Consumer Finances (SCF). We present results from an econometric framework, the Parameterised Dogit model, that allows us to separately, and simultaneously, model individual heterogeneity (i.e. supply side) and employment type heterogeneity (i.e. demand-side) influences that determine self-employment. Our findings suggest that whilst individual characteristics are important determinants of self-employment, there are also factors which are specific to the type of employment that influence whether an individual is self-employed.

An investigation of the impact of promotions on across-submarket competition

There is a recent trend among manufacturers of consumer packaged goods such as Proctor & Gamble to eliminate sales promotions and replace them with everyday low prices on their products. The objective of this study is to investigate whether and how such a significant change in promotional strategy would affect the competitive patterns in a market. Specifically, what would be the impact of such a move on competition between different brands in a market? Would the competitive structure change significantly as a result? How would this affect specific brands? We propose a methodology based on a dogit model of brand switching to evaluate and test the impact of promotions on competitive structure and to identify the nature of the impact. Our findings reveal that some sales promotions (e.g., displays and price cuts) have a significant impact in shaping competitive patterns characterizing submarkets for crackers. The magnitude of this impact decreases with increased product differentiation and increased intensity of the promotional activities in the market. Our findings help managers determine how the elimination of certain sales promotions may reshape competitive patterns before making such a move. The knowledge of which submarket boundaries are stable and which boundaries can be affected by promotions also helps managers decide which product alternatives to promote (or not to promote) so as to maximize short-term revenues.

A Monte Carlo study of tests for the independence of irrelevant alternatives property

A plethora of tests for the Independence of Irrelevant Alternatives (IIA) property of Logit models of discrete choice behavior has been proposed in the literature. These tests are based upon asymptotic arguments and little is known about their size and power properties in finite samples. This paper uses a Monte Carlo simulation study to investigate the size and power properties of six tests for IIA in the multinomial Logit model. Our results show that the majority of tests based upon partitioning the choice set appear to have very poor size and power properties in small samples. Tests for IIA based upon the DOGIT model, similarly have poor size properties, but in some circumstances do have reasonable power properties.

Analytical structure of discrete choice models

In this paper a new conceptual framework of the external interventions of the active socio-economic and territorial environment in the content of the choice between the set of competitive alternatives is considered. This conceptual framework generates a new analytical procedure for the derivation of all possible mode-specific choice models, starting from the preset initial “building block” models. The analytical basis for this procedure is the description of all mappings transforming the multidimensional unit simplex of choice probabilities into other unit simplices. These mappings represent the totality of the redistributions of adopters between competitive alternatives caused by the interventions of an active environment. It is shown, for example, that the choice probabilities for the standard dogit model can be generated by the multiplication of the specific Markovian matrix by the vector of choice probabilities of the logit model. The components of this Markovian matrix are the transition probabilities of the transfer of adopters between alternatives caused by superimposition of the action of environment. Analogously, each nested model corresponds to the product of several Markovian matrices. Starting from the standard logit model, the following mappings were used for the derivation of an array of possible choice models: ratio-linear, linear, log-linear, ratio-polynomial and polynomial mappings. The result was generation of the choice models by McFadden, Gaudry, Gaudry and Dagenais, Borgers and Timmermans and Fotheringham, which are already known from the literature, together with several new choice models.

The dogit model is applicable even without perfectly captive buyers

The dogit choice model is a modification of the logit model to allow for ‘captive’ or ‘perfect loyal’ buyers. This paper derives the dogit choice model from a two-period loyalty model drawn from the marketing science literature. This will prove that the dogit model can be used even when buyers are not perfectly captive (or ‘loyal’) to one choice. We also deduce an equilibrium dogit model corresponding to the equilibrium choice probabilities when buyer choices from period to period follow the two-period marketing science loyalty model.

A Diagnostic Test for the Multinomial Logit Model

I propose a Lagrange multiplier test for the multinomial logit model against the dogit model (Gaudry and Dagenais 1979) as the alternative hypothesis. In view of the well-known drawback of the restrictive property of independence from irrelevant alternatives implied by the multinomial logit model, a specification test has much to recommend it. Finite sample properties of the test are studied using a Monte Carlo experiment, and the test's power against the nested multinomial logit model and the multinomial probit model is investigated. The test is found to be sensitive to the values of the regression parameters of the linear random utility function.

The inverse power transformation logit and dogit mode choice models

This paper explores the properties of inverse Box-Cox and Box-Tukey transformations applied to the exponential functions of logit and dogit mode choice models. It is suggested that inverse power transformations allow for the introduction of modeler ignorance in the models and solve the “thin equal tails” problem of the logit model; it is also shown that they allow for asymmetry of response functions in both logit and dogit models by introducing alternative-specific parameters which make cross elasticities of demand among alternatives generally asymmetric. In the dogit model, modeler ignorance and consumer captivity remain conceptually distinct. Standard logit and dogit models appear as very special “perfect knowledge” cases in broad spectra of models which also include, among others, the reciprocal extreme value or log-Weibull variants. These improvements over the simple symmetric-thin-equal-tail-perfect-knowledge logit and the symmetric-pure-captivity dogit are achieved at the cost of introducing at the most two new parameters per alternative considered in the original logit and dogit mode choice models.

The dogit model

This paper presents the dogit model. That model is flexible enough to permit the choice among specific pairs of alternatives to be consistent with the independence from irrelevant alternatives axiom, as in a logit model, but it simultaneously allows the choice among other pairs not to be. Dogit parameters add an “income effect” to the “substitution effect” already built into the logit model; alternatively, they allow for the joint presence of compulsive and discretionary elements in consumer behavior, or for the identification of captive markets. Eventual estimation of the values of the parameters of the dogit model appears simpler than for the probit model.

Testing the dogit model with aggregate time-series and cross-sectional travel data

This paper compares dogit and logit specifications of market share models, taking into account the possibility that conclusions might depend on transformations of the explanatory variables of these models. Parameter estimates are obtained both for a time-series urban transit mode of payment model and for a cross-sectional intercity mode choice model. It is demonstrated, using current maximum likelihood techniques extended to take multiple-order autocorrelation of the residuals into account, that the dogit specification is at least equal to, and sometimes clearly superior to, the logit specification irrespective of transformations of explanatory variables.