Mixed RP/SP models incorporating interaction effects

In order to analyse the impact of a new train service in Cagliari (Italy) a databank including information from a revealed preference (RP) and a stated preference (SP) survey was set up. The RP data concern choice between car, bus and train; the SP data consider the binary choice between a new train service (quicker, more frequent, with a lower fare and more stations than the current one) and the alternative currently chosen by car and bus users. Logit models allowing for correlation among RP alternatives were estimated for this mixed RP/SP data set using the artificial tree structure method. The analysis included level-of-service variables measured with an unusually high level of precision, latent or second order variables (such as comfort), inertia and interaction variables. Different specifications of the utility function were tested, including the expenditure rate model, and the effects of these specifications on modelling results are highlighted. Our results show that for a population mainly composed of fixed income workers, the expenditure rate model is superior to the traditional wage rate model, yielding lower and more significant subjective values of time. Moreover, we found that the non-linear specifications appear to be more suitable as not only better model results were obtained, but also the real distribution of the error terms was revealed (i.e. highlighting correlation among public transport options).     

Accounting for inertia in modal choices: some new evidence using a RP/SP dataset

Inertia is related with effect that experiences in previous periods may have on the current choice. In particular, it has to do with the tendency to stick with the past choice even when another alternative becomes more appealing. As new situations force individuals to rethink about their choices new preferences may be formed. Thus a learning process begins that relaxes the effect of inertia in the current choice. In this paper we use a mixed dataset of revealed preference (RP)-stated preference (SP) to study the effect of inertia between RP and SP observations and to study if the inertia effect is stable along the SP experiments. Inertia has been studied more extensively with panel datasets, but few investigations have used RP/SP datasets. In this paper we extend previous work in several ways. We test and compare several ways of measuring inertia, including measures that have been proposed for both short and long RP panel datasets. We also explore new measures of inertia to test for the effect of “learning” (in the sense of acquiring experience or getting more familiar with) along the SP experiment and we disentangle this effect from the pure inertia effect. A mixed logit model is used that allows us to account for both systematic and random taste variations in the inertia effect and for correlations among RP and SP observations. Finally we explore the relation between the utility specification (especially in the SP dataset) and the role of inertia in explaining current choices.     

Temporal transferability of updated alternative-specific constants in disaggregate mode choice models

Methods of updating disaggregate discrete choice models have been proposed as a means of obtaining better transferability. However, the temporal transferability of models updated for better spatial transferability has rarely been analysed, and the factors affecting temporal transferability have not been determined. This paper deals with one updating method—the use of disaggregate data to update alternative-specific constants—and investigates the factors affecting the temporal transferability of the updated constants. In the analysis, repeated cross-section data collected in the Chukyo metropolitan area are divided, efficiently generating many application areas. The analysis showed that the factors can depend on regional characteristics and past travel behaviours (inertia), and are anti-symmetric and path-dependent of changes in the level of service.     

Correcting state dependence and serial correlation in the RP/SP combined estimation method

Revealed preference (RP) data and stated preference (SP) data have complementary characteristics for model estimation. To enhance the advantages of both data types, a combined estimation method is proposed. This paper discusses the method and practical considerations in applying it, and introduces a new method of considering serial correlation of RP and SP data. An empirical analysis is also presented.     

A practical concern about the relevance of alternative-specific constants for new alternatives in simple logit models

When a “new” alternative is introduced, post-estimation, into a logit model, analysis initially exclude an ASC or occasionally assume a correspondence with an existing alternative. Using a recent data set for mode and route choice, the paper highlights the forecasting implications of ignoring the ASC in the utility expression of a “new” alternative, and how sensitive the market shares are to the inclusion/ exclusion of the ASC.     

A sequential approach to exploiting the combined strengths of SP and RP data: Application to freight shipper choice

The possibility of and procedure for pooling RP and SP data have been discussed in recent research work. In that literature, the RP data has been viewed as the yardstick against which the SP data must be compared. In this paper we take a fresh look at the two data types. Based on the peculiar strengths and weaknesses of each we propose a new, sequential approach to exploiting the strengths and avoiding the weaknesses of each data source. This approach is based on the premise that SP data, characterized by a well-conditioned design matrix and a less constrained decision environment than the real world, is able to capture respondents' tradeoffs more robustly than is possible in RP data. (This, in turn, results in more robust estimates of share changes due to changes in independent variables.) The RP data, however, represent the current market situation better than the SP data, hence should be used to establish the aggregate equilibrium level represented by the final model. The approachfixes the RP parameters for independent variables at the estimated SP parameters but uses the RP data to establish alternative-specific constants. Simultaneously, the RP data are rescaled to correct for error-in-variables problems in the RP design matrixvis-à- vis the SP design matrix. All specifications tested are Multinomial Logit (MNL) models.The approach is tested with freight shippers' choice of carrier in three major North American cities. It is shown that the proposed sequential approach to using SP and RP data has the same or better predictive power as the model calibrated solely on the RP data (which is the best possible model for that data, in terms of goodness-of-fit figures of merit), when measured in terms of Pearson's Chi-squared ratio and the percent correctly predicted statistic. The sequential approach is also shown to produce predictions with lower error than produced by the more usual method of pooling the RP and SP data.     

On the development of time period and mode choice models for use in large scale modelling forecasting systems

A substantial amount of research is presently being carried out to understand the complexities involved in modelling the choice of departure time and mode of travel. Many of these models tend to be far too complex and far too data intensive to be of use for application in large scale model forecasting systems, where socio-economic detail is limited and detailed scheduling information is rarely available in the model implementation structure. Therefore, these models generally work on the basis of a set of mutually exclusive time periods, rather than making use of continuous departure time information. Two important questions need to be addressed in the use of such models, namely the specification used for the time periods (in terms of length), and the ordering of the levels of nesting, representing the difference in the sensitivities to shifts in departure time and changes in the mode of travel. This paper aims to provide some answers to these two questions on the basis of an extensive analysis making use of three separate Stated Preference (SP) datasets, collected in the United Kingdom and in the Netherlands. In the analysis, it has proved possible to develop models which allow reasonably sound predictions to be made of these choices. With a few exceptions, the results show higher substitution between alternative time periods than between alternative modes. Furthermore, the results show that the degree of substitution between time periods is reduced when making use of a more coarse specification of the time periods. These results are intended for use by practitioners, and form an important part of the evidence base supporting the UK Department for Transport’s advice for practical UK studies in the WebTAG system.¹     

Flexible substitution patterns in models of mode and time of day choice: new evidence from the UK and the Netherlands

Modelling the temporal response of travellers to transport policy interventions has rapidly emerged as a major issue in many practical transport planning studies and is recognised to hold particular challenges. The importance of congestion and its variation over the day, together with the emergence of time-dependent road user charging as a policy tool, emphasise the need to understand whether and how travellers will change the timing of their journeys. For practical planning studies, analysts face a major issue of relating temporal changes to other behavioural changes that are likely to result from policy or exogenous changes. In particular, the relative sensitivity of time and mode switching has been difficult to resolve. This paper describes a study undertaken to determine the relative sensitivity of mode and time of day choice to changes in travel times and costs and to investigate whether evidence exists of varying magnitudes of unobservable influences in time of day switching. The study draws on data from three related stated preference studies undertaken over the past decade in the United Kingdom and the Netherlands and uses error components logit models to investigate the patterns of substitution between mode and time of day alternatives. It is concluded that the magnitude of unobserved influences on time switching depends significantly on the magnitudes of the time switches considered. With time periods of the magnitude generally represented in practical modelling, i.e. peak periods of 2–3 hours, time switching is generally more sensitive in these data than mode switching. However, the context of the modelling and the extent to which relevant variables can be measured will strongly influence these results.     

Analytical models for comparison of alternative service options for the transportation handicapped

Recently much debate has been generated over the issue of public transportation service for the transportation handicapped. In particular, older rapid rail systems have been required to make key stations accessible to the handicapped. However, a waiver of this requirement is permitted in cities where the handicapped community and local officials work out an alternative service system with equal or better station accessibility. In this context, it is the purpose of this paper to present a comparative analysis of a door-to-door demand responsive system and feeder/fixed-route service, two of the service options which are expected to represent acceptable alternatives to “key transit station accessibility” requirements. The comparisons evaluate alternative scenarios of doorstep accessible transportation for the transportation handicapped using analytical models. The results of the comparisons clearly indicate the relative advantage of door-to-door demand-responsive systems over the combination of a feeder and an existing fixed-route system. In particular, when compared to a feeder service, the direct door-to-door service is likely to be less costly, because it requires a smaller fleet and to provide a more desirable service, because it does not entail transfers.     

Treatment of seasonal tickets for public transport in estimation and application of mode/destination choice models

A new model system dealing with trips of length up to 100 km has recently been developed in Norway. A new way of dealing with seasonal passes for public transport is used in the travel-to-work model. The objective was to account for the fact that a respondent that posses a seasonal pass for public transport may behave as if public transport is free on the day they report a travel diary. On the other hand, we can not assume that public transport is free for respondents that used other modes of transport or that public transport is free to alternative destinations. This problem was solved by defining seasonal pass as a separate alternative in the form of a nest that included all modes of travel. The cost of a seasonal pass is a common cost for all modes in the nest and will thus not affect the choice within the nest. The estimation of this specification is compared with the more common approach of assigning an average cost per day based on the cost of a monthly pass and the number of workdays in a month. The comparison indicates that the “average cost per day” approach may produce biased estimates for several parameters. It also turns out that the cost parameter for seasonal pass is higher than the parameter for “out of pocket” cost, probably reflecting that there will be some uncertainty with respect to the actual use of a seasonal pass.     

Additive models for network data envelopment analysis in the presence of shared resources

This paper studies the problem of evaluating the relative efficiency of a set of specialized and interdependent decision-making subunits that make up a large decision-making unit (DMU). The paper develops a data envelopment analysis (DEA) approach for measuring the efficiency of decision processes which can be divided into two stages. In these processes the first stage uses its own inputs to generate outputs which a part of these outputs become the inputs to the second stage. Moreover, DMUs use shared input sources in both operation stages. This paper provides a set of additive models which measure the performance of two-stage network DEA processes with shared inputs. Numerical examples show the applicability of the approach.     

An investigation on the performances of mode shift models in transit ridership forecasting

This paper aims at investigating the over-prediction of public transit ridership by traditional mode choice models estimated using revealed preference data. Five different types of models are estimated and analysed, namely a traditional Revealed Preference (RP) data-based mode choice model, a hybrid mode choice model with a latent variable, a Stated Preference (SP) data-based mode switching model, a joint RP/SP mode switching model, and a hybrid mode switching model with a latent variable. A comparison of the RP data-based mode choice model with the mode choice models including a latent variable showed that the inclusion of behavioural factors (especially habit formation) significantly improved the models. The SP data-based mode switching models elucidated the reasons why traditional models tend to over-predict transit ridership by revealing the role played by different transit level-of-service attributes and their relative importance to mode switching decisions. The results showed that traditional attributes (e.g. travel cost and time) are of lower importance to mode switching behaviour than behavioural factors (e.g. habit formation towards car driving) and other transit service design attributes (e.g. crowding level, number of transfers, and schedule delays). The findings of this study provide general guidelines for developing a variety of transit ridership forecasting models depending on the availability of data and the experience of the planner.     

Development of commuter and non-commuter mode choice models for the assessment of new public transport infrastructure projects: A case study

This paper uses state of the art stated choice designs to parameterise modal choice models for commuting and non-commuting travel futures in the presence of new public transport infrastructure (variations of new heavy rail, light rail and dedicated busway systems). D-optimal choice experiments are developed for a set of labelled modal alternatives in which respondents establish a reference benchmark based on the existing service levels (for access, linehaul and egress trip legs) which is used in a computer aided personal interview instrument to generate future scenarios of service levels for current and prospective new modals options. We show that a fully integrated stated choice experiment provides all the information required to obtain behaviourally relevant parameter estimates (within a nested logit framework) for all but the mode-specific constants (MSCs). The MSCs can be calibrated for the current modes within a network model setting, giving the transport planner an appropriate model for predicting the patronage potential for proposed new public transport infrastructure services. A useful by-product is a new set of behavioural values of travel time savings for access, egress, linehaul and wait times.     

An investigation of prediction errors in work trip mode choice models

This paper reports some analyses of predictive accuracy of disaggregate work trip mode choice models. The prediction error is separated into three components: model specification error, aggregation error, and transfer error. The main results are (1) the total forecasting errors can be very large, especially if the model transfers poorly; (2) poor transfer-ability was found between cities where the transit or shared ride markets are much different; (3) prediction from models having statistically different coefficients need not be much different and (4) the type of level-of-service data, i.e. manually coded or network based, used in model estimation has little influence on forecasting accuracy. These and other results are provisional for reasons discussed in the main text.     

Effects of incorporating latent and attitudinal information in mode choice models

Abstract

Hybrid choice modelling approaches allow latent variables in mode choice utility functions to be addressed. However, defining attitude and behavior as latent variables is influenced by the researcher's assumptions. Therefore, it is better to capture the effects of latent behavioral and attitudinal factors as latent variables than defining behaviors and attitudes per se. This article uses a hybrid choice model for capturing such latent effects, which will herein be referred to as modal captivity effects in commuting mode choice. Latent modal captivity refers to the unobserved and apparently unexplained attraction towards a specific mode of transportation that is resulting from latent attitude and behavior of passengers in addition to the urban transportation system. In empirical models, the latent modal captivity variables are explained as functions of different observed variables. Empirical models show significant improvement in fitting observed data as well as improved understanding of travel behavior.     

The sensitivity of parameter estimates to data specification in mode choice models

Standard network data are generally used in estimation of mode choice models. These data are inaccurate in several ways, but the cost of correcting the inaccuracies is great. This paper analyzes the effects which correcting some of the inaccuracies in the standard network data has on the estimated parameters of mode choice models. Models are estimated on the standard network data and on data which have been adjusted so as to correct the problems in the standard network data. It is found that, for analysis of policies affecting transfer wait times or distances to bus stops, correction of the standard network data is advisable. For other policy analyses, however, it seems that the extra expense of correcting the standard data is unnecessary.     

A comparison of the predictive ability of mode choice models with various levels of complexity

Models of mode choice have recently been developed which include a large number of explanatory variables. The inclusion of some of these variables is obviously the result of trial-and-error analysis of various model specifications: the researcher tries various specifications until he obtains a specification which is consistent with a priori beliefs and fits the data fairly well. This method of model specification allows one to “learn” from the data, but is also open to the critism that the resultant model simply reflects relations which happen to exist in the sample, rather than true, behavioral relations.This paper examines this question. A complex model is presented which was developed after attempting a wide variety of specifications. The predictive ability of this model is compared with that of models with fewer variables, each of which could be included on the basis of a priori ideas. It is found that the complex model predicts best, indicating that the behavioral content of the model which was developed through “learning” from the data is greater than that of models which were specified on a priori beliefs.     

Comparison of updating techniques in transferability analysis of work trip mode choice models in developing countries

This study analyzes the performances of updating techniques in transferability of mode choice models in developing countries. A model specification, estimated in Ho Chi Minh City, was transferred to Phnom Penh. Naïve transfer and four updating methods associated with small sized samples were used in the transfer process and were evaluated based on statistical perspective and predictive ability. The study also illustrates the problems faced in model transferability development, due to the lack of available and suitable data in Phnom Penh. This lack is strongly related to different methods and structures applied in collecting the data. Simplified approaches to the difficulties are proposed in the study. The results show that updating ASCs, updating both ASCs and scale parameter, and use of combined transfer estimators all produce significant improvement, both statistically and in predictability, in updating the model. The last two methods have proven to be superior to the first method, owing to the inclusion of transfer bias considerations in the estimations. However, small data samples should not have large transfer bias when using combined transfer estimators. It is also concluded that naïvely transferring a model is not recommended, and Bayesian updating should be avoided when transfer bias exists. Copyright © 2010 John Wiley & Sons, Ltd.