Attribute Range Effects in Binary Response Tasks

This paper investigates attribute range effects in binary response conjoint analysis tasks. We investigate a long-standing conjecture that the regression estimates of attributes in choice tasks are influenced by a researcher's selected range of attribute levels across choice sets. Specifically, we examine the effect(s) of varying attribute ranges systematically over two ranges of levels (1=wide range, 2=levels in the middle of the wide range) in a public bus choice context. A master 2³ design is used to vary the range (i.e., wide, middle) of three numerical attributes (fare, service and time). In each of the eight master range conditions a 2³ factorial creates bus profiles, and a ninth condition is added to test for response non-linearity. Our results suggest that attribute range impacts attribute main effects to a small degree, yet exhibit substantial and systematic effects on attribute interactions and model goodness-of-fit. Implications of these results for practical design of academic and commercial choice-based conjoint analysis tasks are discussed.     

Heterogeneity distributions of willingness-to-pay in choice models

We investigate direct and indirect specification of the distribution of consumer willingness-to-pay (WTP) for changes in product attributes in a choice setting. Typically, choice models identify WTP for an attribute as a ratio of the estimated attribute and price coefficients. Previous research in marketing and economics has discussed the problems with allowing for random coefficients on both attribute and price, especially when the distribution of the price coefficient has mass near zero. These problems can be avoided by combining a parameterization of the likelihood function that directly identifies WTP with a normal prior for WTP. We show that the typical likelihood parameterization in combination with what are regarded as standard heterogeneity distributions for attribute and price coefficients results in poorly behaved posterior WTP distributions, especially in small sample settings. The implied prior for WTP readily allows for substantial mass in the tails of the distribution and extreme individual-level estimates of WTP. We also demonstrate the sensitivity of profit maximizing prices to parameterization and priors for WTP.

Dissecting the Random Component of Utility

We illustrate and discuss several general issues associated with the random component of utility, or more generally unobserved variability. We posit a general conceptual framework that suggests a variance components view as an appropriate structure for unobserved variability. This framework suggests that unobserved heterogeneity is only one component of unobserved variability; hence, a more general view is required. We review a considerable amount of empirical research that suggests that random components are unlikely to be independent of systematic components, and random component variances are unlikely to be constant between or within individuals, time periods, locations, etc. We also review evidence that random components are functions of (elements of) systematic components. The latter suggests considerable caution in the use and interpretation of complex choice model specifications, in particular recently introduced forms of random parameter models that purport to estimate distributions of preference parameters. Several areas for future research are identified and discussed.     

A dimensional versus attribute approach for disaggregate choice models

A fundamental issue facing choice modelers is to make a decision on what kind of independent variables to include in a choice model. With survey data, the two immediate options are: actual product attributes or underlying latent dimensions (factor scores). Using behavioral logic we argue that heterogeneity of consumer perceptions of variables and their saliences should be the key items moderating such a decision. We present empirical evidence to support our theory that dimensional (factor score) based models do better in terms of predictions than attribute based models in more heterogeneous populations. Empirical analysis shows that in segments (where consumer heterogeneity is lower) the predictive performance of attribute based models improves relative to the factor score model and may actually have a better predictive fit when the respondents are relatively homogeneous with respect to attribute ratings and saliences.     

Multi-stage purchase decision models: Accommodating response heterogeneity, common demand shocks, and endogeneity using disaggregate data

The most comprehensive models of purchase behavior for frequently purchased supermarket items explain households' purchase incidence decisions (whether to buy), brand choice decisions (what to buy), and purchase quantity decisions (how much to buy). In this study, we develop a three-stage purchase incidence/brand choice/purchase quantity model for household-level data in which all three stages are specified with (i) random coefficient distributions for model covariates and (ii) random effect distributions to account for unobserved factors affecting demand (known as common demand shocks), while also (iii) controlling for the effects of endogeneity on prices. Compared to current state-of-the-art models for multi-stage purchase decisions, the results show improvements in fit and forecasting accuracy when purchase behaviors are modeled with all of these components in combination. Perhaps more importantly, when common demand shocks are ignored, substantial differences in parameter estimates and diagnostic information about consumer behavior are likely (median differences in parameter estimates are 10% and 20% in two product categories), which impact managerial deliberations about price and promotion policies. Further, failure to account for common demand shocks affects the mean and variance of random coefficient distributions in unpredictable directions, which could produce results that encourage managers to pursue inappropriate and costly micro-level product marketing strategies.     

Using a Community of Knowledge to Build Intelligent Agents

The modeling of individual consumer preference can be aided by incorporating others' opinions which contain information above and beyond identified product attributes. The value of others' opinions is tested using two empirical data sets. The results indicate that incorporating others' opinions into an attribute-based model can reduce systematic error and increase predictive accuracy by serving as a proxy for missing information (e.g., undiscovered attributes or attribute interactions, sensory or experiential aspects of the product, as well as advertising or word of mouth effects). Additionally, modeling individual preference based on others' opinions alone is shown to predict as well or better than traditional multiattribute models thus bypassing the need for defining a product attribute space.     

The effects of traffic light labels and involvement on consumer choices for food and financial products

Traffic light (TL) labels that inform consumers regarding product safety have received increasing attention in different fields. Behind the background of behavioural economics, this paper presents the results of a split‐sample choice experiment conducted in Germany to evaluate the impact of TL labelling on purchases of food and financial products. We hypothesize that consumers experience different levels of involvement with these two types of products, leading to different recommendations regarding the use of TLs. The results show that TLs affect consumer purchases of both product groups by focusing their attention on specific product attributes. For food, whereas the low‐fat attribute has no significant impact on food choices that do not include TLs, this attribute has a positive impact on choices once it is signalled with a TL label. The positive evaluation of the organic production attribute of a food product without a TL decreases when the same product is labelled with a TL. In the case of financial products, TLs significantly reinforce the impact of all characteristics on choice probability. TLs also generate a halo effect with regard to return variance. Although consumers demonstrate different levels of involvement for the two types of products, involvement does not always impact the evaluation of attributes. Compared with less involved consumers, more involved consumers exhibit more heterogeneous evaluations of the return variability attribute in the case of the financial product and the organic attribute in the case of the food product.     

Shaping online consumer choice by partitioning the Web

This research explores how partitioning attributes in online search interfaces changes the valuations of those attributes—and impacts subsequent choice—such that attributes that are displayed as separate categories tend to receive greater decision weight than attributes grouped under umbrella categories. Across several choice domains—cars, dates, and hotels—we show that different attribute partitions impact the importance assigned to attributes (Studies 1 and 2), as well as consumer choices (Studies 3 and 4). We argue that these effects are due in part to users' willingness to use the implicit recommendations of interface designers to determine the importance of attributes, a willingness that extends to following explicit recommendations of online agents based on those attributes (Study 5). © 2009 Wiley Periodicals, Inc.     

Understanding consumers’ multichannel choices across the different stages of the buying process

This article provides a more integrative approach toward channel choice than previous research by considering all stages of the buying process (search, purchase, and after-sales), and by taking channel attributes, experience, and spillover effects into account when examining consumers?? channel choice intentions. The authors show that such an integrative perspective is important as channel attributes, experience, and spillover matter for consumers?? channel choices in all stages of the buying process. Notably, the study stresses the importance of channel experience and spillover effects for explaining consumers?? channel choice intentions in the different stages of the buying process. Channel experience effects occur when using the channel increases the likelihood that the consumer will use the very same channel on the next occasion. Spillover effects result when the likelihood of using a channel in one stage of the buying process affects the likelihood of choosing that channel in another stage. The results show that both effects influence consumers?? channel choice intentions over and above channel attributes. Importantly, the model results strongly pledge for studying attribute, experience, and spillover effects simultaneously.     

Attribute Identities Matter: Subjective Perceptions of Attribute Characteristics

Recent research indicates that attributes vary along multiple dimensions with implications for how trade-offs are resolved during choice. We present an exploratory study of the dimensionality underlying naïve subjects ratings of attributes on the characteristics commonly discussed in the literature on tradeoff resolution and decision difficulty. Factor analysis of attribute characteristic assessments indicates that subjects view decision attributes in a multi-dimensional fashion, including an importance/loss aversion dimension, an emotional potential/protection from tradeoffs dimension, and a cognitive difficulty dimension. These results suggest that a one-dimensional measure of attribute characteristics, such as a standard attribute importance rating, may obscure some factors determining individual responses to attributes during decision processing. However, the results also suggest that developing a relatively succinct set of scales in order to characterize the dimensions along with subjects response to attributes is a viable goal for future research.     

Three kinds of order effects in choice-based conjoint analysis

Question and response order effects are known to occur in survey research instruments. Choice-based conjoint analysis presents respondents with multiple sets of concepts fully specified in terms of a number of attributes. Respondents choose one full-profile concept from each of the many choice sets. Thus there are three plausible order effects with respect to choice-based conjoint analysis: choice set order, profile order within choice sets, and attribute order within profiles. Three empirical studies show that statistically significant choice set order, profile order, and attribute order effects occur. Some of these order effects are significant from a practical standpoint, but none occur in a predictable way. Analysts are advised to rotate, across respondents, profile and attribute order to offset the order biases. Directions for future research are suggested.     

Sensitivity of multiattribute decision models to alternative specifications of weighting parameters

The issue of attribute weighting in multiattribute decision models is examined. Results are presented which show that the outputs produced by linear multiattribute models are extremely robust with respect to alternative specifications of the weighting parameters unless the number of attributes included in the models is small, the average correlation among the attributes is low, and the dispersion of the weights is large relative to their mean. Implications of these results are discussed for three different types of weighting schemes-regression weighting, equal weighting, and subjective weighting—which are used in multiattribute decision modeling.     

Choice deferral: The interaction effects of visual boundaries and consumer knowledge

Choice deferral is a common and lousy phenomenon for companies and consumers. This research examines how visual boundaries (by-alternatives vs. attributes) interact with consumer knowledge to influence choice deferral. Five studies (N = 1,245) and two single-paper meta-analyses demonstrate that by-alternative (vs. attribute) visual boundaries reduce choice deferral and that choice difficulty mediates this effect. The research also shows that these effects are stronger for less knowledgeable consumers. However, the interactive effect of visual boundaries and consumer knowledge on choice deferral depends on task and context characteristics. We discuss future research about choice deferral moderators and how to facilitate consumer choice.     

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.     

Best–Worst Scaling Approach to Predict Customer Choice for 3PL Services

This study describes a simple, theoretically based methodology to analyze the nature of customer demand for third‐party logistics provider service components. The method overcomes limitations in prior studies and enables us to examine the relative importance of product and service attributes as they pertain to the choice of third‐party logistics providers. Two distinct types of customers populate our data: those professing operational attributes and those seeking relational attributes. The theoretical and practical implications are that improved supply chain models can be developed when separate demand structures are taken into account.     

Enhancing marketing with engineering: Optimal product line design for heterogeneous markets

Successful product line design and development often require a balance of technical and market tradeoffs. Quantitative methods for optimizing product attribute levels using preference elicitation (e.g., conjoint) data are useful for many product types. However, products with substantial engineering content involve critical tradeoffs in the ability to achieve those desired attribute levels. Technical tradeoffs in product design must be made with an eye toward market consequences, particularly when heterogeneous market preferences make differentiation and strategic positioning critical to capturing a range of market segments and avoiding cannibalization.We present a unified methodology for product line optimization that coordinates positioning and design models to achieve realizable firm-level optima. The approach overcomes several shortcomings of prior product line optimization models by incorporating a general Bayesian account of consumer preference heterogeneity, managing attributes over a continuous domain to alleviate issues of combinatorial complexity, and avoiding solutions that are impossible to realize. The method is demonstrated for a line of dial-readout scales, using physical models and conjoint-based consumer choice data. The results show that the optimal number of products in the line is not necessarily equal to the number of market segments, that an optimal single product for a heterogeneous market differs from that for a homogeneous one, and that the representational form for consumer heterogeneity has a substantial impact on the design and profitability of the resulting optimal product line — even for the design of a single product. The method is managerially valuable because it yields product line solutions efficiently, accounting for marketing-based preference heterogeneity as well as engineering-based constraints with which product attributes can be realized.     

Multiple Attribute Group Decision-Making Methods with Completely Unknown Weights in Intuitionistic Fuzzy Setting and Interval-Valued Intuitionistic Fuzzy Setting

The existing multiple attribute group decision-making approaches based on intuitionistic fuzzy sets (IFSs) or interval-valued intuitionistic fuzzy sets (IVIFSs) are considered as the situation that the weights of experts are given beforehand and the attribute weights are known or unknown. To better describe the uncertain decision environment and solve the corresponding decision problem, multiple attribute group decision-making methods with completely unknown weights of both experts and attributes are proposed in intuitionistic fuzzy setting and interval-valued intuitionistic fuzzy setting. Entropy weight models can be used to determine the weights of both experts and attributes from intuitionistic fuzzy decision matrices or interval-valued intuitionistic fuzzy decision matrices, and then the evaluation formulas of weighted correlation coefficients between alternatives and the ideal alternative are introduced in intuitionistic fuzzy setting and interval-valued intuitionistic fuzzy setting. The alternatives can be ranked and the most desirable one(s) can be selected according to the values of the weighted correlation coefficients for IFSs or IVIFSs. Finally, two numerical examples demonstrate the effectiveness of the proposed methods: they are capable for handling the multiple attribute group decision-making problems with completely unknown weights.     

A hybrid conjoint model for the identification of the UK supermarket choice determinants: an exploratory study

Supermarket competition has increased tremendously during recent years and the need for a deeper understanding of the key criteria for consumers' store patronage has never been greater. Against this background, the present study tries to identify and analyse the basic determinants that drive supermarket choice. This paper, which is primarily methodological, applies a large-scale hybrid conjoint experiment to a UK setting with the purpose of examining the relative impact of various supermarket attributes on the actual choice. The hybrid conjoint model deals with the problem of designing and analysing studies involving large numbers of attributes and/or attribute levels, and makes it possible to estimate individually based part-worth values, determines the most important supermarket choice criteria and yields useful implications for strategic retail management.     

A comparison of importance weights and willingness-to-pay measures derived from choice-based conjoint, constant sum scales and best–worst scaling

We review the measurement of product attribute importance, and find little consensus in definition or measurement methods. We compare four measurement methods: 1) two direct methods whereby respondents report the importance of attributes using best–worst scaling or constant sum scales, and 2) two indirect methods derived from discrete choice experiments. Our comparisons rely on previous findings that choice experiments are externally valid to use as the standard. We find high agreement within direct or indirect methods, but less agreement between direct and indirect methods. Our results also demonstrate that inferences derived from indirect measures appear to be susceptible to context effects related to the particular attributes a researcher chooses to investigate. We discuss implications for current and future research.