Random regret minimization for consumer choice modeling: Assessment of empirical evidence

This paper introduces to the field of marketing a regret-based discrete choice model for the analysis of multi-attribute consumer choices from multinomial choice sets. This random regret minimization (RRM) model, which has recently been introduced in the field of transport, forms a regret-based counterpart of the canonical random utility maximization (RUM) paradigm. This paper assesses empirical results based on 43 comparisons reported in peer-reviewed journal articles and book chapters, with the aim of finding out to what extent, when, and how RRM can form a viable addition to the consumer choice modeler's toolkit. The paper shows that RRM and hybrid RRM–RUM models outperform RUM counterparts in a majority of cases, in terms of model fit and predictive ability. Although differences in performance are quite small, the two paradigms often result in markedly different managerial implications due to considerable differences in, for example, market share forecasts.     

An Analysis of Assortment Choice in Grocery Retailing

Consumers in grocery retailing commonly buy bundles of products to accommodate current and future consumption. When all products in a particular bundle share common attributes (and are selected from the same product category), the consumer is said to assemble an assortment. This research examines the impact of assortment variety on the assortment choice process. In particular, we test the prediction that consumers demand less variety for higher quality items. To investigate this relationship, we employ a flexible choice model, suitable for the analysis of assortment choice. The model, based upon the assumption that the utility of purchase of one item in an assortment depends upon the set of items already selected, allows for a general utility structure across the assortment items. We apply the model to household assortment choice histories from the yogurt product category. Substantively, we show that yogurt choice is affected by brand quality perceptions (quality-tier competition). Moreover, we show that reaction to reductions in variety (number of yogurt flavors) is mediated by brand quality perceptions. Taken together, these empirical facts paint a picture of a consumer who is willing to trade-off variety against product quality in assortment choice.     

A New Look at Constructed Choice Processes

Normative models of choice assert axiomatically that preferences are consistent, coherent, and determined only by relevant alternatives. In contrast to this classical economic perspective, behavioral models derived from research in psychology and consumer behavior assert that preferences are not guided by an internal, stable utility function but are constructed during the choice process. The current paper is based on a session on constructed choice processes (CCP) at the 2004 Choice Symposium that focused on how the standard CCP model can be enriched by bringing theories and tools from modern research in social cognition to bear on choice phenomenon. The richer conceptual framework presented by new, currently unpublished empirical work provides a novel perspective on choice construction by integrating the roles of subjective construal, experiential information, attribution, goals, and satisfaction in understanding preference construction processes in choice.     

The Effect of Attribute Variation on Consumer Choice Consistency

We study the effect of shifts in attribute level differences on consumer choice consistency. Choice consistency is measured as the variance of the random error component in the consumer utility function: the smaller this variance, the higher choice consistency. We hypothesize that due to increased choice difficulty, choice consistency decreases if attribute level differences increase while average utility level differences between alternatives remain the same. In our empirical illustration we focus on the impact of price level shifts on choice consistency in conjoint choice experiments. Our results show that choice consistency decreases as price level differences increase and absolute price levels increase.     

Cumulative Discrete Choice

We present a discrete choice model in which a consumer's impression of each alternative is based on her memory of past experience with this choice, and is stochastically updated whenever the alternative is chosen. The consumer remembers a cumulative utility index per alternative, and, when an alternative is chosen, the index is updated by the addition of a random variable, interpreted as instantaneous utility. We prove that the frequencies of choice converge, with probability 1, to limit frequencies, which can be computed from the model's parameters.     

Modeling Methods for Discrete Choice Analysis

This paper introduces new forms, sampling and estimation approaches fordiscrete choice models. The new models include behavioral specifications oflatent class choice models, multinomial probit, hybrid logit, andnon-parametric methods. Recent contributions also include new specializedchoice based sample designs that permit greater efficiency in datacollection. Finally, the paper describes recent developments in the use ofsimulation methods for model estimation. These developments are designed toallow the applications of discrete choice models to a wider variety ofdiscrete choice problems.     

Modeling preference evolution in discrete choice models: A Bayesian state-space approach

We develop discrete choice models that account for parameter driven preference dynamics. Choice model parameters may change over time because of shifting market conditions or due to changes in attribute levels over time or because of consumer learning. In this paper we show how such preference evolution can be modeled using hierarchial Bayesian state space models of discrete choice. The main feature of our approach is that it allows for the simultaneous incorporation of multiple sources of preference and choice dynamics. We show how the state space approach can include state dependence, unobserved heterogeneity, and more importantly, temporal variability in preferences using a correlated sequence of population distributions. The proposed model is very general and nests commonly used choice models in the literature as special cases. We use Markov chain monte carlo methods for estimating model parameters and apply our methodology to a scanner data set containing household brand choices over an eight-year period. Our analysis indicates that preferences exhibit significant variation over the time-span of the data and that incorporating time-variation in parameters is crucial for appropriate inferences regarding the magnitude and evolution of choice elasticities. We also find that models that ignore time variation in parameters can yield misleading inferences about the impact of causal variables. This paper is based on the first author's doctoral dissertation.     

Representing Heterogeneity in Consumer Response Models 1996 Choice Conference Participants

We define sources of heterogeneity in consumer utility functions relatedto individual differences in response tendencies, drivers of utility, formof the consumer utility function, perceptions of attributes, statedependencies, and stochasticity. A variety of alternative modelingapproaches are reviewed that accommodate subsets of these various sourcesincluding clusterwise regression, latent structure models, compounddistributions, random coefficients models, etc. We conclude by defining anumber of promising research areas in this field.     

Decision Strategy and Structure in Households: A “Groups” Perspective

From a practical perspective, (arguably) most consumer decisions are not made in isolation of the households in which consumers are inserted, yet we commonly treat them econometrically as if they were. The purpose of this workshop was to take some initial steps in defining needed research in household decision making that structurally accounts for goal sharing, utility interdependence, taste heterogeneity, choice set formation, power structures, group size and composition, and so forth. We also considered conditions under which aggregation of tastes, utility and choices might occur and make sense from both behavioral and modeling perspectives.     

The market value for product attribute improvements under price personalization

Personalization of the marketing mix is a topic of much interest to marketing academics and practitioners. Using discrete choice demand theory, we investigate the aggregate market value for product attribute improvements when firms are engaged in personalized pricing. Our results provide a theoretically grounded rule for how to aggregate consumer valuations to assess the overall profitability of attribute improvements under price personalization. Under common pricing, each consumer contributes the same margin. Profitability of an attribute improvement is thus driven by inducing more consumers to buy. Consumers with high choice probabilities are given less weight in the market valuation under common pricing as they are less responsive to attribute improvements. Under personalized pricing, profitability of an attribute improvement is driven by extraction of consumer surplus from high valuation consumers. Consumers with higher valuations, and consequently higher choice probabilities, are given more weight in the market valuation under personalized pricing. Since individual consumers play a more central role in the market valuation under personalized pricing, estimation of consumer-level valuations is of increased importance. Under common pricing, the market valuation for an attribute improvement is robust to extreme estimates of the consumer-level valuations. Through our theoretical and empirical analyses, we demonstrate that this robustness does not hold under personalized pricing.     

Recent Progress on Endogeneity in Choice Modeling

We describe recent progress in several areas related to endogeneity, including: choice set formation and attention to attributes; interactions among decision-makers; respondents' strategic behavior in answering stated preference choices; models of multiple discrete/continuous choice; distributions of willingness-to-pay; and methods for handling traditionally endogenous explanatory variables.     

The use of multinomial choice analysis in international business research

Many strategic decisions made by firms involve a choice among several discrete alternatives. International Business (IB) scholars are often interested in modelling the factors that potentially influence these (multinomial) choices: these factors might include not only characteristics of the firms making the choices but also attributes of the alternative choices. This paper provides a succinct and intuitive introduction to the possible applications of multinomial choice models in IB research. We briefly outline the theory behind discrete choice modelling, and then explain how multinomial choice models may be estimated (including how the datasets need to be formatted) and how the significance of the coefficient estimates as well as the diagnostic statistics may be tested and interpreted.     

Accounting for heterogeneity in shopping centre choice models

Random utility models are a tool of great interest in the study of shopping centre choice. However, most research that has used these models employed the most basic specification: the multinomial logit model. This model presents two main drawbacks. Firstly, it is assumed that sensitivity to variables of attraction and dissuasion can be considered the same for all consumers. Secondly, this model can only be applied to situations in which alternatives from which you can choose are totally independent. In this paper, we present two specifications which allow us both to introduce heterogeneity and relax the assumption of independence of irrelevant alternatives, a nested logit model and a random effects model. Data collected in three European cities of the same region are used to implement these models. Results attained from these models reveal the existence of different segments of consumers. This provides the potential for retail managers to gain better insight into the way in which different consumers choose between a set of shopping centres in a certain area.     

Where to refuel: Modeling on-the-way choice of convenience outlet

For many goods consumers do not make a special trip to a store. Especially for a convenience good such as fuel they will buy the product while on-the-way to some final destination. This paper introduces on-the-way choice of retail outlet as a form of convenience shopping. It presents a model of on-the-way choice of retail outlet and applies the model in the context of fuel retailing to explore its implications for segmentation and spatial competition. The model allows analyzing how choice of retail outlet varies not only with spatio-temporal variables (distance, detour, local competition and agglomeration) but also with trip-related characteristics such as time of day and prior awareness of one's purchase need. The model is a latent class random utility choice model. An application to gas station choices observed in a medium-sized Asian city show the model to fit substantially better than existing models. The empirical results indicate consumers may adopt one of two decision strategies. When adopting an immediacy-oriented strategy they behave in accordance with the traditional gravity-based retail models and tend to choose the most spatially convenient outlet. When following a destination-oriented strategy they focus more on maintaining their overall trip efficiency and so will tend to visit outlets located closer to their main destination and are more susceptible to retail agglomeration effects. The paper demonstrates how the model can be used to inform segmentation and local competition analyses that account for variations in these strategies as well as variations in consumer type, origin and time of travel. Simulations of a duopoly setting further demonstrate the implications.     

The role of self selection, usage uncertainty and learning in the demand for local telephone service

Telephone services are often characterized by the presence of ‘fixed’ plans, involving only a fixed monthly fee, as well as ‘measured’ plans, with both fixed fees and per-unit charges for usage. Consumers are faced with the decisions of which plan to choose and how much to use the phone and these decisions are not, in general, independent. Due to the presence of a time lag between plan choice and usage decisions, consumers are uncertain about usage at the plan-choice stage. We develop a structural discrete/continuous model of plan choice and usage decisions of consumers that accounts for such uncertainty. Prior research has also found that consumers switch less often from fixed plans to measured plans to gain from potential savings than vice versa. Consumer uncertainty regarding their mean usage levels and different rates of learning by consumers in the two plans is a potential explanation for this phenomenon. We extend our discrete/continuous model to account for consumer learning about their mean usage and estimate different rates of learning for the two types of plans. We estimate our model using data from the 1986 Kentucky local telephone tariff experiment. Even in the absence of any price variation over time, we are able to measure the price elasticities both of usage and of choice of plan. Using our parameter estimates, we simulate the effects of the introduction of a metered plan in a market with only a fixed plan and vice versa, on both firm revenues and consumer surplus. We also find that consumers learn very rapidly if they are on the measured plan but learn very slowly when they are on the fixed plan. We investigate an alternative assumption on the nature of the learning process in which only consumers in the measured plan have an opportunity to learn. We find that our empirical results are robust to this change of specification. We conduct counterfactual simulations to simulate enhanced calling plans from the firm and consumer points of view. Additional simulations to measure the value of information in this category are also carried out. We compute the value of both complete information, where the entire uncertainty about future usage is resolved, as well as that of limited information, where the consumer's uncertainty about mean usage is resolved, but the uncertainty about specific month-to-month usage remains. We find that the value of information is modest. We also find that a large proportion of the value of information is that about the mean usage, with the value of the information about a specific month's usage being relatively small.

When self‐customization backfires: The role of a maximizing mindset

Self‐customization in online shopping contexts readily offers an abundance of options for consumers. However, the sheer amount of information can quickly become overwhelming. One way to give people the freedom to choose without overwhelming them with information is to simplify the decision process by breaking it down into a series of smaller steps. Contrary to a common assumption that simpler decisions increase choice likelihood, however, this study demonstrates that a simple by‐attribute self‐customization process may activate a maximizing mindset, which increases people's desire to find better options and decreases their satisfaction with the ones available. Consequently, simplifying the self‐customization process can sometimes backfire by decreasing choice likelihood. Three studies suggest that although by‐attribute self‐customization formats are easier to choose from, compared with more complex matrix formats, they may sometimes—paradoxically—increase choice deferral. The findings suggest that a maximizing mindset mediates this effect, casting doubt on information‐based alternative explanations. The findings also suggest that whether by‐attribute self‐customization increases or decreases choice likelihood may depend on the presence of objective quality cues, which indicate that an objectively‐best option can be found. This study furthers the understanding of how decision difficulty and maximizing influence self‐customization decisions.     

A simple method for estimating preference parameters for individuals

This paper demonstrates a method for estimating logit choice models for small sample data, including single individuals, that is computationally simpler and relies on weaker prior distributional assumptions compared to hierarchical Bayes estimation. Using Monte Carlo simulations and online discrete choice experiments, we show how this method is particularly well suited to estimating values of choice model parameters from small sample choice data, thus opening this area to the application of choice modeling. For larger sample sizes of approximately 100–200 respondents, preference distribution recovery is similar to hierarchical Bayes estimation of mixed logit models for the examples we demonstrate. We discuss three approaches for specifying the conjugate priors required for the method: specifying priors based on existing or projected market shares of products, specifying a flat prior on the choice alternatives in a discrete choice experiment, or adopting an empirical Bayes approach where the prior choice probabilities are taken to be the average choice probabilities observed in a discrete choice experiment. We show that for small sample data, the relative weighting of the prior during estimation is an important consideration, and we present an automated method for selecting the weight based on a predictive scoring rule.     

Competing Online: Studying Consumer Choice Drivers for Online Financial Services

ABSTRACT

Given the rapid increase in the consumer use of online services and the increase in competition between firms that compete online, firms are faced with a crucial challenge. Having invested significant resources in transitioning consumers from using offline services to using online services, they now need to understand what drives consumers to choose between competing online services. Our study seeks an exploratory answer to the above challenge. Specifically, we consider, “what role do factors that drive consumers into using online services play in assisting firms better compete in the online space?” This paper explores the above question by quantifying the value that consumers of an online financial service place on having access to in-depth product information, an affordable online service, an easy to use online service, access to offline capabilities, and available marketing promotions. The results reported in this paper are based on a web-based discrete choice experiment in which 2,209 consumers were asked to compare various online financial service offerings, differing from each other in terms of the relative availability of our critical factors. The results demonstrate that consumer preferences (relative utilities) for various factors of an online financial service are different. Our results enable practicing managers to understand the factors that drive consumer choice when faced with competing online services. We believe that these results have both managerial and research implications for design, management and operations strategy formulation for online services.     

Perspectives on Multiple Category Choice

Multiple category choice is a decision process in which an individualselects a number of goods, all of which are nonsubstitutable with respect toconsumption. Choices can be made either simultaneously or sequentially. Thekey feature of multiple category choice is the treatment of the choices asinterrelated because each item in the final collection of goods contributesto the achievement of a common behavioral goal. We discuss current andpotential applications of psychology, economics and consumer choice theoryin developing models of multiple category choice.     

Learning About Computers: An Analysis of Information Search and Technology Choice

We estimate a dynamic model of how consumers learn about and choose between different brands of personal computers (PCs). To estimate the model, we use a panel data set that contains the search and purchase behavior of a set of consumers who were in the market for a PC. The data includes the information sources visited each period, search durations, as well as measures of price expectations and stated attitudes toward the alternatives during the search process. Our model extends recent work on estimation of Bayesian learning models of consumer choice behavior in environments characterized by uncertainty by estimating a model of active learning—i.e., a model in which consumers make optimal sequential decisions about how much information to gather prior to making a purchase. Also, following the suggestion of Manski (2003), we use our data on price expectations to model consumers’ price expectation process, and, following the suggestion of McFadden (1989a), we incorporate the stated brand quality information into our likelihood function, rather than modeling only revealed preference data.Our analysis sheds light on how consumer forward-looking price expectations and the process of learning about quality influence the consumer choice process. A key finding is that estimates of dynamic price elasticities of demand exceed estimates that ignore the expectations effect by roughly 50%. This occurs because our estimated expectations formation process implies that consumers expect mean reversion in price changes. This enhances the impact of a temporary price cut. Finally, while our work focuses specifically on the PC market, the modeling approach we develop here may be useful for studying a wide range of high-tech, high-involvement durable goods markets where active learning is important.JEL Classification: C15, C33, C35, C42, C51, C52, D83, D84