Journal of Choice Modelling最新文献

Advanced econometric models used in the field of transport or marketing are becoming increasingly sophisticated and able to capture complex decision making and outcomes. In this paper, we apply state-of-the-art discrete-continuous choice models to the field of Ecology, in particular to model activity engagement of the population of Swedish Brown bears. Using data from GPS collars that track wild animals over time, we estimate a Multiple Discrete-Continuous Extreme Value (MDCEV) model to understand activity engagement and duration as a function of both bear characteristics and other external factors. We show that the methodology is not only suitable to address this aim, but also allows us to produce insights into the connection between the animal's age and gender and activity engagement as well as the links with climate variables (temperature and precipitation) and human activity (hunting).

Many multivariate model systems involve the estimation of a covariance matrix that must be positive-definite. A common strategy to ensure positive definiteness of the covariance matrix is through the use of a Cholesky parameterization of the covariance matrix. However, several model systems require imposing restrictions on the elements of the covariance elements. For instance, modelling systems may require fixing some (or all) of the diagonal elements in the covariance matrix to unity due to identification considerations. However, imposing such restrictions using the traditional Cholesky decomposition approach is not feasible and requires the additional parameterization of the Cholesky elements.

In this paper, we explore a separation-based strategy with spherical parameterization of the Cholesky matrix to impose restrictions on the covariance matrix. Importantly, using this separation-based parameterization strategy, we also explore the possibility of restricting some covariance (or correlation) terms to zero. The effectiveness of the proposed strategy is assessed through extensive simulation experiments. The results from the simulation experiments highlight better performance of the separation-based strategy in terms of recovery of model parameters – particularly those in the covariance matrix, than the traditional Cholesky parameterization approach. Finally, the proposed strategy is implemented in a joint multivariate binary probit ordered probit model system to analyze the usage (and the extent of use) of non-private modes of transportation in Bengaluru, India. In doing so, the proposed strategy is implemented to restrict several correlations to zero, thus avoiding the estimation of a profligate correlation matrix and substantially easing the estimation process.

Sales of deer licenses, one of the most important revenue sources for wildlife management at the Indiana Department of Natural Resources (IDNR), have been declining for a decade. To increase its revenue, the IDNR is considering introducing a new lifetime deer license for sale. This license would allow hunters to harvest deer (and possibly other species) each year for the rest of their lives in exchange for a relatively large up-front fee. The forward-looking nature of the decision to buy a lifetime license means hunters' choice behavior is necessarily dynamic. Prior work estimates preferences for long-lived, durable goods using standard discrete choice experiments underpinned by static models. We derive a dynamic discrete choice model of lifetime license purchases. Our model informs the design of a novel, dynamic discrete choice experiment, generating data that allows us to consistently estimate individuals’ forward-looking preferences for lifetime hunting licenses. We use our model to estimate the price of lifetime licenses that maximizes IDNR revenues.

Estimation of discrete outcome specifications involves significant hypothesis testing, including multiple modelling decisions which could affect results and interpretation. Model development is generally time-bound, and decisions largely rely on experience, knowledge of the problem context and statistics. There is often a risk of adopting restricted specifications, which could preclude important insights and valuable behavioral patterns. This study proposes a framework to assist in testing hypotheses and discovering mixed-Logit specifications that best capture discrete outcome behavior. The proposed framework includes a mathematical programming formulation and a bi-level constrained optimization algorithm to simultaneously test various modelling assumptions and produce meaningful specifications within a reasonable time. The bi-level framework illustrates the integration of a population-based metaheuristic with model estimation procedures. In addition, the optimization algorithm allows the analyst to impose assumptions on the models to test specific hypotheses or to ensure compliance with literature. Numerical experiments are conducted using different datasets and behavioral processes to illustrate the efficacy of the proposed extensive hypothesis testing in terms of interpretability and goodness-of-fit. Results illustrate the ability of the proposed algorithm to reveal important insights that can potentially be overlooked due to limited and/or biased hypothesis testing. In addition, the proposed extensive hypothesis testing generates multiple acceptable solutions, thereby suggesting potential directions for further investigation. The proposed framework can serve as a decision-assistance modelling tool in various applications, involving many variables and outcomes, such as road safety analysis, consumer choice behavior, and integrated land-use and travel choice models.

When estimating random coefficients models from choice data, decisions relating to the multivariate density function assumed to describe preference heterogeneity across the population raise questions about stochastic (in)dependence between preference dimensions, uni- vs. multi-modality, potential point masses, bounds and/or constraints on support regions, among other concerns. Parametric representations of population distributions have generally implied uncomfortable compromises to achieve estimation tractability. It would seem preferable to sidestep such issues by estimating individual preferences in a distribution-free manner, but this freedom of form implies a large number of parameters since we lose the parsimony enabled by parametric densities and must deal directly with estimation of individual decision maker preferences. I propose a hybrid distribution-free estimator for individual parameter logit models that uses a genetic algorithm as first stage, the solution from which becomes a starting point for a gradient-based search to obtain the final posterior maximum likelihood estimates of individual preferences. This estimator is described in detail, its parameter recovery capability is tested with Monte Carlo data generation simulations, and a case study is developed in some detail to illustrate its use in policy analysis. The estimator can be applied to both stated and revealed preference data, requiring only that sufficient choice replications be available for individual observation units consistent with extant estimation methods. Computational experience shows the estimator to require CPU times comparable to extant simulation-based estimation methods, meaning that its use is practical for the exploration of the parameter space through multiple trials.

Revealed and stated choice data are fundamental inputs to understanding individuals’ preferences. Owning to the distinctive characteristics and complementary nature of these two types of data, making joint inference based on their combined information content represents an attractive approach to preference studies. However, complications may arise from the different decision protocols under the two distinct choice contexts. In this study, a Bayesian hierarchical model is proposed to make joint inference from combined RP and SP data, with special attention paid to capturing the behavioural differences between the two choice contexts. In addition to the well-recognised issues of decision inertia and scale differences, the proposed model also takes into account other behavioural characteristics such as a decision-maker ignoring situation constraints, non-attending attributes, and misinterpreting attributes. An empirical analysis of a combined RP and SP dataset of travel mode choices is used to demonstrate the advantageous features of the model. Upon examining the empirical evidence, two main advantages emerge: the model provides direct measures of the effect of behavioural issues arising from ignoring situation constraints and non-attending attributes, as well as evidence for the misinterpretation of attributes.

Taking a holiday trip is a common couple-based leisure activity in which both partners tend to be actively involved. This paper studies the intra-household bargaining for the choice of a vacation destination within couples. We conduct a discrete choice experiment in which we elicit both individual and couple preferences for different hypothetical travel portfolios in a two-stage experimental design. The couple choices are modelled as a function of males' and females' individual preferences, allowing for different bargaining weights for each characteristic of the holiday trip. Therefore, we assess partners’ bargaining power (influence) in the couple choices conditional on individual preferences. We find that although males have a more influential role overall, there seems to be a gender specialization in that females decide on the type of accommodation and males focus on the trip cost.

Individual migration mobilities over the life course have not been well understood in existing studies, and therefore ways to represent the underlying intertemporal dynamics and heterogeneities have remained unclear. To fill this research gap, this study investigates the domestic migration of people residing in the Capital Area of Japan, which has suffered from various issues caused by the over-concentration of population for several decades. Using a web-based questionnaire survey, workers aged 20–49 living in the Capital Area were requested to recall their five latest migration experiences (i.e., migration biography). A life-course intertemporal discrete choice model with cross-sectional and longitudinal heterogeneities was developed to represent individual migration destination biographies, by introducing quasi-hyperbolic utility and drawing on time preference theory. It was found that a considerable proportion of working people in the Capital Area (especially Tokyo) are from other regions of Japan. In the modeling analysis, the temporally-changing, intertwined and heterogeneous roles of place attachment, motives and altruism in migration decisions over the life course are empirically confirmed. Nonlinear influences of past, present and future utilities are further revealed, where the past utility grows more influential, and the importance of future utility diminishes over time. Policy implications of the derived findings for the development of megacities and local cities are discussed.

An important component of the design phase of a discrete choice experiment (DCE) is formulating the cost vector, which specifies the costs of the alternatives and enables the calculation of marginal willingness to pay (WTP) estimates. If the cost vector affects choice behaviour, welfare estimates may depend on the choice of the cost vector, which leads to problems with the validity and reliability of DCE results. We employ a split-sample design to examine cost vector effects on choice behaviour and WTP estimates. Our data come from a DCE on agri-environmental policies to a nationally representative sample in Finland. We provide additional insights compared to previous research by including four cost vectors with otherwise identical surveys and experimental designs and a positive cost for the status quo alternative, with cost levels for policy alternatives both below and above the status quo cost. We obtain some evidence that the cost vector affects choice behaviour, as the proportion of status quo choices is larger with higher cost vectors. Both absolute and relative cost levels matter for choices. The marginal WTP estimates are highest in the sub-sample with the largest range cost vector that has cost levels both below and above the status quo cost. We suggest more careful pre-testing of the cost levels compared to current practices to determine a plausible range of cost levels to produce valid welfare estimates.

We propose three procedures based on association rules (AR) learning and random forests (RF) to support the specification of a portfolio choice model applied in data from complex choice experiment data, specifically a Participatory Value Evaluation (PVE) choice experiment. In a PVE choice experiment, respondents choose a combination of alternatives, subject to a resource constraint. We combine a methodological-iterative (MI) procedure with AR learning and RF models to support the specification of parameters of a portfolio choice model. Additionally, we use RF model predictions to contrast the validity of the behavioural assumptions of different specifications of the portfolio choice model. We use data of a PVE choice experiment conducted to elicit the preferences of Dutch citizens for lifting COVID-19 measures. Our results show model fit and interpretation improvements in the portfolio choice model, compared with conventional model specifications. Additionally, we provide guidelines on the use of outcomes from AR learning and RF models from a choice modelling perspective.