Random Choice Model Research Articles

Integrating random regret minimization-based discrete choice models with mixed integer linear programming for revenue optimization

Integrating random regret minimization-based discrete choice models with mixed integer linear programming for revenue optimization

Consumption and preferences for mushrooms in Ghana: A comparison between regret-based and utility-based approaches

This paper employs descriptive survey and advanced discrete choice modelling approaches to examine consumption behaviour and preferences for mushrooms in Ghana. Specifically, the paper compares the mixed regret-based and utility-based approaches to examine consumer choices. Using a choice experiment data, the findings revealed, on average, that consumers hold preferences for straw mushroom variety and a disutility for the shiitake mushroom variety, but they are indifferent towards the oyster mushroom variety. The findings further indicated that consumers prefer less expensive and locally-cultivated mushrooms to expensive and imported mushrooms. Methodologically, the findings show that the mixed random regret-based choice models outperform the mixed random utility-based models. Nonetheless, the results are comparable. It is recommended that government’s policy be directed towards promoting cost-effective straw and oyster mushroom production to sustain the industry.

Eliciting farmers’ preferences towards agriculture education in Northern Ireland

ABSTRACT Purpose Agriculture education promotes farmers and societal welfare through a more efficient use of resources and the adoption of sustainable farming practices. Nevertheless, farmers’ educational attainment is limited across countries. This study explores farmers’ intention to participate in formal agriculture education. Design/methodology/approach A discrete choice experiment survey with 363 farmers exploring their preferences for different specifications within a diploma in agriculture was employed. We focus on Northern Ireland, a country with a relatively large agricultural sector and were farmers have a low level of formal education. Findings Our results show that farmers exhibited a preference for short courses delivered in the evenings, with 75% of the teaching time delivered online, favouring educational topics linked to farm performance against those related to socio-environmental outcomes. Part-time younger farmers, operating in small farms located in less favoured areas, reported the highest willingness to pay for the course. However, farmers that had previously participated in a Young Farmer Payment Scheme reported the lowest interest in undertaking the course. Practical implications The study provides useful information to education providers in relation to the design of agriculture courses to better meet farmers’ preferences and therefore encourage levels of uptake. Secondly, the study can help policymakers to make informed decisions about policy instruments incentivising farmers to engage in education. Theoretical implications By drawing from random parameter choice models, the study highlights that the heterogeneity of farmers’ preferences should be considered when modelling their decision to engage in agriculture education. Originality/value This is the first study that elicits farmers’ preferences towards agriculture education via a discrete choice experiment. Secondly, the study explores preference heterogeneity based on farmers and farms’ characteristics.

The Focal Luce Model

We introduce the Focal Luce Model (FLM), a random choice model that generalizes Luce’s (1959) model (multinomial logit) to account for menu-dependent focality of alternatives. In the FLM, focal alternatives are relatively more likely to be chosen even after taking utilities into account. The FLM captures a variety of bounded rationality models while behaviorally distinguishing between what is focal and the magnitude of the bias from focality. We show how to identify utilities, focal sets, and bias terms from choice frequencies; behaviorally characterize the FLM by weakening Luce’s (1959) independence from irrelevant alternatives; and apply our model to choice data and games. (JEL C70, D11, D91)

Dynamic Labor Reallocation with Heterogeneous Skills and Uninsured Idiosyncratic Risk

Occupational specificity of human capital motivates an important role of occupational reallocation for the economy’s response to shocks and for the dynamics of inequality. We introduce occupational mobility, through a random choice model with dynamic value function optimization, into a multi-sector/multi-occupation Bewley (1980)- Aiyagari (1994) model with heterogeneous income risk, liquid and illiquid assets, price adjustment costs, and in which households differ by their occupation-specific skills. Labor income is a combination of endogenous occupational wages and idiosyncratic shock. Occupational reallocation and its impact on the economy depend on the transferability of workers’ skills across occupations and occupational specialization of the production function. The model matches well the statistics on income and wealth inequality, and the patterns of occupational mobility. It provides a laboratory for studying the short- and long-run effects of occupational shocks, automation and task encroaching on income and wealth inequality. We apply the model to the pandemic recession by adding an SIR block with occupation-specific infection risk and a ZLB policy and study the impact of occupational and aggregate labor supply shocks. We find that occupational mobility may tame the effect of the shocks but amplifies earnings inequality, as compared to a model without mobility.

Kısmi Rasyonel Bireyler İçin Basit Bir Seçim Modeli

We analyze a restricted pro-con random choice model of Dogan and Yildiz (2018) in which there are only two observable orderings relevant for choice such as price and quality. An agent is endowed with two orderings that both serve as pro-orderings and con-orderings. For each choice set, if an alternative is the top-ranked (bottom-ranked) by one of these orderings, then this is a pro (con) for choosing that alternative. Each pro or con item has a weight reflecting its salience. Each alternative is chosen with a probability proportional to the total weight of its pros and cons. In our analysis, we provide a set of choice axioms, which guarantee that the observed choices can be generated via a pro-con random choice model with respect to two given observable orderings. A notable aspect of this characterization is that the associated axioms discipline the choice behavior when choice sets present attraction effect scenarios. This indicates the tight connection between the model and the attraction effect phenomena.

Dynamic Labor Reallocation with Heterogeneous Skills and Uninsured Idiosyncratic Risk

Occupational specificity of human capital motivates an important role of occupational reallocation for the economy's response to shocks and for the dynamics of inequality. We introduce occupational mobility, through a random choice model with dynamic value function optimization, into a multi-sector/multi-occupation Bewley-Aiyagari model with heterogeneous income risk, liquid and illiquid assets, price adjustment costs, and in which households differ by their occupation-specific skills. Labor income is a combination of endogenous occupational wages and idiosyncratic shock. Occupational reallocation and its impact on the economy depend on the transferability of workers' skills across occupations and occupational specialization of the production function. The model matches well the statistics on income and wealth inequality, and the patterns of occupational mobility. It provides a laboratory for studying the short- and long-run effects of occupational shocks, automation and task encroaching on income and wealth inequality. We apply the model to the pandemic recession by adding an SIR block with occupation-specific infection risk and a ZLB policy and study the impact of occupational and aggregate labor supply shocks. We find that occupational mobility may tame the effect of the shocks but amplifies earnings inequality, as compared to a model without mobility.

What Is the Impact of Nonrandomness on Random Choice Models?

Problem definition: This paper examines the impact of nonrandomness on random choice models and studies various operations problems under the new discrete choice models. Academic/practical relevance: The literature often assumes that the random utility components follow some independent and identically distributed distribution. This assumption is too restrictive in some real-world scenarios, because, for example, consumers may have known well about the attribute values for the product that they have repeatedly purchased. Methodology: We adopt the random utility maximization framework and characterize the choice probabilities when the utility of some alternative is deterministic. The log-likelihood function is jointly concave in the attribute coefficients under the linear utility-attribute assumption; an expectation-maximization algorithm is developed to overcome the missing data issue in estimation. Results: Surprisingly, if the utility of a particular product is deterministic, the assortment problem is still polynomial-time solvable, whereas if the utility of the no-purchase option is deterministic, the decision problem corresponding to the assortment optimization is NP-complete. We show that the price minus the reciprocal of price sensitivity is product invariant at optimality, which helps to simplify the multiproduct pricing problems. Managerial implications: Empirical study on real data shows that incorporating nonrandomness into random choice models can increase model fitting and prediction accuracy. Failure of accounting for the impact of nonrandomness may result in substantial losses.

Odds supermodularity and the Luce rule

We present a characterization of the Luce rule in terms of positivity and a new choice axiom called odds supermodularity that strengthens the regularity axiom. This new characterization illuminates a connection that goes unnoticed, and sheds light on the behavioral underpinnings of the Luce rule and its extensions from a different perspective. We show that odds supermodularity per se characterizes a structured extension of the Luce rule that accommodates zero probability choices. We identify the random choice model characterized via a stochastic counterpart of Plott (1973)'s path independence axiom, which strengthens odds supermodularity.

The Decision-Conflict and Multicriteria Logit

We study two random choice models that explain behavioural patterns suggesting that the outside option is often chosen by decision makers who find it hard to decide between the otherwise desirable market alternatives available to them. The decision-conflict logit extends the logit model with an outside option by assigning a menu-dependent value to that option that allows for its choice probability to either increase or decrease when the menu is expanded, depending on which as well as on how many alternatives are added to it. The multicriteria logit is a special case of this model that introduces multiple utility functions which jointly predict behaviour in a novel multiplicative-logit way. This special case admits a general discrete-choice formulation.

What is the Impact of Non-Randomness on Random Choice Models?

Problem definition. This paper examines the impact of non-randomness on random choice models, and then studies various operations problems under the new discrete choice models. Academic/Practical Relevance. The literature often assumes that the random utility components follow some i.i.d. distribution. This assumption may be too restrictive in some real-world scenarios, because for example some consumers may know well about attribute values for some product (which could also be the no-purchase option) that they have repeatedly purchased. Methodology. We adopt the random utility maximization framework, and characterize the choice probabilities when the utility of some alternative is deterministic. The log-likelihood function is still jointly concave; the EM algorithm is developed to overcome the missing data issue. Results. Surprisingly, if the utility of a particular product is deterministic, the assortment problem is polynomial-time solvable; whereas if the utility of the no-purchase option is deterministic, the assortment problem is NP-hard. We show that the prices are product-invariant at optimality and use this result to simplify the multi-product pricing problems. Managerial Implications. Empirical study on real data shows that incorporating non-randomness into random choice models can increase model fitting and prediction accuracy. Failure of accounting for the impact of non-randomness on random choice models may result in substantial losses.

Error distributions assumptions in random regret choice models: towards error Frechit specifications

Recently introduced regret-based choice models in transportation research have mainly adopted the assumption of identically and independently distributed unobserved regrets. The central argument underlying this paper is that this assumption is difficult to defend considering the fundamental nature of the concept of regret, which states that regret is generated through the comparison of choice alternatives on an attribute-by-attribute basis. To support this stance, we identify and diagnose specification errors in classic regret-based choice models, and provide an alternative specification. Results show that relaxing the assumption of identically and independently distributed error terms and applying the proposed error components Frechit structure leads to a substantial improvement in model performance of the regret-based models for the data used.

Luce Rule with Attraction System

We develop an extension of Luce's random choice model that incorporates a role for the association of alternatives. Each alternative is characterized by a salience value, a Luce value, and its associated alternatives. The salience value captures the alternative's ability to attract the decision maker's attention, and the Luce value measures the alternative's desirability. The decision maker is first attracted by some alternative according to a salience-based Luce-type formula, and then chooses among its associated alternatives according to another desirability-based Luce-type formula. While retaining the simplicity of the Luce rule, the theory accommodates some well-known behavioral phenomena in individual choice, such as the attraction effect (violations of regularity), and violations of stochastic transitivity.

Research on influence of sun glare in urban tunnels based on cellular automaton model in the framework of Kerner’s three-phase traffic theory

As ventilation shafts set at the top of urban tunnels becomes a common pattern for ventilation, the sunlight cast under the shafts could cause sun glares which negatively affect the driver’s visual performance. In this paper, the influence of sun glare on individual drivers in urban tunnels is quantitatively illustrated by the change rate of driver’s pupillary area. Then random choice model of deceleration behavior in response to sun glare (DBSG) is established. The DBSG in urban tunnels is incorporated into the well-known KKSW (Kerner–Klenov–Schreckeneberg–Wolf) CA model and a new cellular automata model, KKSW-UT (urban tunnels) CA model is proposed. By means of the numerical simulation, traffic flow characteristics in urban tunnel are studied in the framework of Kerner’s three-phase traffic theory. And the frequency and severity of rear-end conflicts in urban tunnels are analyzed to study the influence of DBSG on traffic safety. Based on the results, it is found that DBSG can negatively influence both the traffic operation and safety.

A New Characterization of the Luce Rule With an Extension to Zero Probability Choices

We present a characterization of the Luce rule in terms of positivity and a new choice axiom called odds supermodularity that strengthens the regularity axiom. This new characterization illuminates a connection that goes unnoticed and sheds light on the behavioral underpinnings of the Luce rule and its extensions from a different perspective. We show that odds supermodularity per se characterizes a structured extension of the Luce rule that accommodates zero probability choices. We identify the random choice model characterized via a stochastic counterpart of Plott (1973)’s path independence axiom, which strengthens odds supermodularity.

Trading Complex Risks

Complex risks differ from simple risks in that agents facing them only possess imperfect information about the underlying objective probabilities. This paper studies how complex risks are priced by and shared among heterogeneous investors in a Walrasian market. I apply decision theory under ambiguity to derive robust predictions regarding the trading of complex risks in the absence of aggregate uncertainty. I test these predictions in the laboratory. The experimental data provides strong evidence for theory’s predicted reduction in subjects’ price sensitivity under complex risks. While complexity induces more noise in individual trading decisions, market outcomes remain theory-consistent. This striking feature can be reconciled with a random choice model, where the bounds on rationality are reinforced by complexity. When moving from simple to complex risks, equilibrium prices become more sensitive whereas risk allocations turn less sensitive to noise introduced by imperfectly rational subjects. Markets’ effectiveness in aggregating beliefs about complex risks is determined by the trade-off between reduced price sensitivity and reinforced bounded rationality. Moreover, my results imply that complexity has similar but more pronounced effects on market outcomes than ambiguity induced by conventional Ellsberg urns.

The Imbalanced Luce Model

We develop a random choice model in which a decision maker divides the alternatives she faces into two groups, where one group is focal and thus she is more likely to choose alternatives in that group relative to alternatives in the non-focal group: The Focal Luce Model (FLM). The FLM generalizes Luce’s model and naturally captures bounded rationality models while behaviorally distinguishing between what is focal and the magnitude of the bias due to focality. We show how to identify utilities, focal sets, and bias terms from choice frequencies and behaviorally found the FLM with two weakenings of independence from irrelevant alternatives (IIA) which account for the decision maker’s revealed focus. We introduce a new condition on focal sets, Conditionally Decreasing, which captures an increasing difficulty of focusing on multiple alternative and encompasses several models of limited consideration. We apply our model to experimental data, illustrate the importance of accounting for focus in demand estimation, and propose a simple method to identify focal sets.

Extending the logit-mixed logit model for a combination of random and fixed parameters

The logit-mixed logit (LML) model, which allows the analyst to semi-parametrically specify the mixing distribution of preference heterogeneity, is a very recent advancement in logit-type choice models. In addition to generalize many previous semi- and non-parametric specifications, LML is computationally very efficient due to a computationally-convenient likelihood equation that does not require computation of choice probabilities in iterative optimization. However, the original LML formulation assumes all utility parameters to be random. This study extends LML to a combination of fixed and random parameters (LML-FR), and motivates such combination in random parameter choice models in general. We further show that the likelihood of the LML-FR specification loses its special properties, leading to a much higher estimation time. In an empirical application about preferences for alternative fuel vehicles in China, estimation time increased by a factor of 20–40 when introducing fixed parameters. Despite losses in computation efficiency, we show that the flexibility of LML-FR is essential for retrieving eventual multimodality of mixing distributions.

An Expectation-Maximization Algorithm to Estimate the Integrated Choice and Latent Variable Model

As computing capability has grown dramatically, the transport choice model has rigorously included latent variables. However, integrated latent and choice variable (ICLV) models are hampered by a serious problem that is caused by the maximum simulated likelihood method. The method cannot properly reproduce the true coefficients, which is a problem that is often referred to as a lack of empirical identification. In particular, the problem is exacerbated particularly when an ICLV model is calibrated based on cross-sectional data. An expectation-maximization (EM) algorithm has been successfully employed to calibrate a random coefficient choice model, but it has never been applied to the calibration of an ICLV model. In this study, an EM algorithm was adapted to calibrate an ICLV model, and it successfully reproduced the true coefficients in the model. The main contribution of adopting an EM algorithm was to simplify the calibration procedure by decomposing the procedure into three well known econometric problems: a weighted linear regression, a weighted discrete choice problem, and a weighted ordinal choice problem. Simulation experiments also confirmed that an EM algorithm is a stable method for averting the problem of lack of empirical identification.

Bias in random regret models due to measurement error: formal and empirical comparison with random utility model

ABSTRACTThis study addresses the so-called uncertainty problem due to measurement error in random utility and random regret choice models. Based on formal analysis and empirical comparison, we provide new insights about the uncertainty problem in discrete choice modeling. First, we formally show how measurement error affects the random regret model differently from the random utility model. Then, random measurement error is introduced into level-of-service variables and the effect of measurement error is analyzed by comparing the estimated parameters of the concerned choice models, before and after introducing measurement error. We argue that although measurement error leads to biased estimation results in both types of models, uncertainty tends to accumulate in random regret models because this model involves a comparison of alternatives. Therefore, input uncertainty tends to lead to larger bias in random regret models. Moreover, since random regret models assume semi-compensatory decision processes, bias in random utility models is homogenous across individuals and alternatives, while bias in random regret models is heterogeneous. Several approaches are discussed to overcome this uncertainty problem in random regret models.