Search results

The purpose of this paper is to compare and validate the results of two clustering methods for the segmentation of the market for prestige cosmetics in Korea, and to draw conclusions about their general practical usability.

Segmentation schemes based on scaled data collected by questionnaire from 480 female shoppers, selected by the mall intercept procedure, were analysed for validity, model fit, definability of profiles, and usability of results.

Segmentation by traditional K‐means clustering was not judged useful, whereas segments generated by the innovative alternative of mixture regression modelling had clear marketing strategy potential.

Given the single‐country and single‐market context of the study, its outcomes and implications must be generalised cautiously.

Mixture regression can make a significant contribution to the implementation of segmentation strategies based on deliverable consumer benefits, by helping academics and practitioners to better understand, explain and predict patterns of consumer behaviour.

A segmentation model with proven validity offers a sound basis for such marketing strategies as, for example, positioning.

The purpose of this paper is to critically review the latest approaches for capturing and explaining heterogeneity in partial least squares (PLS) path modelling and to classify these into a methodological taxonomy. Furthermore, several areas for future research effort are introduced in order to stimulate ongoing development in this important research field.

Different approaches to treat heterogeneity in PLS path models are introduced, critically evaluated and classified into a methodological taxonomy. Future research directions are derived from a comparison of benefits and limitations of the procedures.

The review reveals that finite mixture‐PLS can be regarded as the most comprehensive and commonly used procedure for capturing heterogeneity within a PLS path modelling framework. However, further research is necessary to explore the capabilities and limitations of the approach.

Directions for additional research, common to most latent class detection procedures include the verification and comparison of available approaches, the handling of large data sets, the allowance of varying structures of path models, the profiling of segments and the problem of model selection.

Whereas modelling heterogeneity in covariance structure analysis has been studied for several years, research interest has only recently been devoted to the question of clustering in PLS path modelling. This is the first contribution which critically consolidates available approaches, discloses problematic aspects and addresses significant areas for future research.

The aim of this paper is to examine product formulation screening at the industrial level in terms of multi‐trait improvement by considering several pertinent controlling factors.

The study adopts Taguchi's orthogonal arrays (OAs) for sufficient and economical sampling in a mixture problem. Robustness of testing data is instilled in this method by employing a two‐stage analysis where controlling components are investigated together while the slack variable is tested independently. Multi‐responses collapse to a single master response has been incurred according to the Super Ranking concept. Order statistics are employed to provide statistical significance. The slack variable influence is tested by regression and nonparametric correlation.

Synergy among Taguchi methodology, super ranking and nonparametric testing was seamless to offer practical resolution to product component activeness. The concurrent modulation of two key product traits due to five constituents in the industrial production of muffin‐cake is invoked. The slack variable, rich cream, is strongly active while the influence of added amount of water is barely evident.

The method presented is suitable only for situations where industrial mixtures are investigated. The case study demonstrates prediction capabilities up to quadratic effects for five nominated effects. However, the statistical processor selected here may be adapted to any number of factor settings dictated by the OA sampling plan.

By using a case study from food engineering, the industrial production of a muffin‐cake is examined focusing on a total of five controlling mixture components and two responses. This demonstration emphasizes the dramatic savings in time and effort that are gained by the proposed method due to reduction of experimental effort while gaining on analysis robustness.

This work interconnects Taguchi methodology with powerful nonparametric tests of Kruskal‐Wallis for the difficult problem of non‐linear analysis of mixtures for saturated, unreplicated fractional factorial designs in search of multi‐factor activeness in multi‐response cases employing simple and practical tools.

Ranked preference data arise when a set of judges rank, in order of their preference, a set of objects. Such data arise in preferential voting systems and market research surveys. Covariate data associated with the judges are also often recorded. Such covariate data should be used in conjunction with preference data when drawing inferences about judges.

To cluster a population of judges, the population is modeled as a collection of homogeneous groups. The Plackett-Luce model for ranked data is employed to model a judge's ranked preferences within a group. A mixture of Plackett- Luce models is employed to model the population of judges, where each component in the mixture represents a group of judges.

Mixture of experts models provide a framework in which covariates are included in mixture models. Covariates are included through the mixing proportions and the component density parameters. A mixture of experts model for ranked preference data is developed by combining a mixture of experts model and a mixture of Plackett-Luce models. Particular attention is given to the manner in which covariates enter the model. The mixing proportions and group specific parameters are potentially dependent on covariates. Model selection procedures are employed to choose optimal models.

Model parameters are estimated via the ‘EMM algorithm’, a hybrid of the expectation–maximization and the minorization–maximization algorithms. Examples are provided through a menu survey and through Irish election data. Results indicate mixture modeling using covariates is insightful when examining a population of judges who express preferences.

Alongside structural equation modeling (SEM), the complementary technique of partial least squares (PLS) path modeling helps researchers understand relations among sets of observed variables. Like SEM, PLS began with an assumption of homogeneity – one population and one model – but has developed techniques for modeling data from heterogeneous populations, consistent with a marketing emphasis on segmentation. Heterogeneity can be expressed through interactions and nonlinear terms. Additionally, researchers can use multiple group analysis and latent class methods. This chapter reviews these techniques for modeling heterogeneous data in PLS, and illustrates key developments in finite mixture modeling in PLS using the SmartPLS 2.0 package.

This research examines the differential impact of the importance of internally and externally sourced information and knowledge and their relationship to absorptive capacity and firm performance. In addition, this analysis deals directly with the unobservable heterogeneity amongst firms that is generally viewed as the raison d'être for a unique resource-based perspective of organizational performance. Latent class, finite mixture regression models are used that show that a single model relating knowledge sourcing, absorptive capacity and firm performance is inadequate in explaining even a minor portion of the variation which is seen between firms.

Since the publication of Van Raaij and Verhallen’s seminal work in European Journal of Marketing in 1994, identifying the domain-specific market segmentation approach as one of the most feasible for segmenting markets, there has been surprisingly limited development in this field, with the food domain as the only exception. This study aims to develop a methodological approach using latent class mixture modelling as contribution in the domain-specific market segmentation field.

This study captures the AIO lifestyle perspective using a domain-specific 80-item algorithm which has the wine (product) domain as its focus. A sample size of 811 consumers is used from data collected by means of the CATI approach.

The authors use four criteria for model selection: comparison of the Bayesian information criterion (BIC) statistic, comparison of classification error, verification of the interpretation of the derived segments and, finally, use of the conditional bootstrap procedure to test whether the selected model provides a significant improvement over the previous model. The five-segment model option yields a minimum BIC, the classification error measure is minimal and is easier to interpret than the other models. Segment descriptions for the five identified lifestyle-based segments are developed.

Segmentation by traditional k-means clustering has proven to be less useful than the more innovative alternative of mixture regression modelling; therefore, the authors identify segments in the market on the basis of individuals’ domain-specific lifestyle characteristics using a latent class mixture modelling approach.

Following the attainment of a clear and robust market segmentation structure, the simultaneous analysis of the lifestyles, demographics and behaviours of consumers as nexus of the domain-specific segmentation approach, provides rich and valid information accurately informing the market segment descriptions.

The authors make a substantive contribution by developing a methodological approach using latent class mixture modelling; the first of its kind in the area of domain-specific segmentation. Next, they use the discriminant and/or predictive validity of the 80-scale items to predict cluster membership using the WRL algorithm. Finally, the authors describe the identified market segments in detail and outline the practical implications.

The pandemic impacted consumers' shopping processes, leading them to approach the online channel for grocery shopping for the first time. The paper contributes to the retailing literature by identifying different grocery shopper segments willing to switch online moved by heterogeneous motivations. Integrating the technology acceptance model 2 (TAM-2) and the protection motivation theory (PMT), this study identifies technology-related and Covid-related motivations jointly impacting channel switching.

A mixture regression model was estimated on the 370 valid questionnaires, filled out by Italian shoppers, delivering four internally consistent segments.

The results reveal the existence of four segments willing to switch towards the online channel for grocery shopping in the aftermath of the pandemic. Utilitarian shoppers would switch online as they consider the online channel useful and easy to use. Responsive shoppers will prefer the online channel driven by the fear of being infected in-store. Novel enthusiasts show interest in the online channel to not catch the virus and cope with emotional fear, although they consider online shopping as an enjoyable and useful activity as well. Smart shoppers consider online shopping as an easy-to-use alternative for their grocery purchases.

This paper identifies technology-related and Covid-related motivations jointly impacting shoppers' channel switching to online and presents a novel method – i.e. mixture regression – allowing for the identification of shopper segments motivated by different reasons, both emotional and utilitarian, to switch towards the online channel for their grocery shopping. Among other motivations, the fear of Covid-19 is identified as a relevant motivation to switch to online.

The purpose of this paper is to provide an overview of unobserved heterogeneity in the context of partial least squares structural equation modeling (PLS-SEM), its prevalence and challenges for social science researchers. Part II – in the next issue (European Business Review, Vol. 28 No. 2) – presents a case study, which illustrates how to identify and treat unobserved heterogeneity in PLS-SEM using the finite mixture PLS (FIMIX-PLS) module in the SmartPLS 3 software.

The paper merges literatures from various disciplines, such as management information systems, marketing and statistics, to present a state-of-the-art review of FIMIX-PLS. Based on this review, the paper offers guidelines on how to apply the technique to specific research problems.

FIMIX-PLS offers a means to identify and treat unobserved heterogeneity in PLS-SEM and is particularly useful for determining the number of segments to extract from the data. In the latter respect, prior applications of FIMIX-PLS restricted their focus to a very limited set of criteria, but future studies should broaden the scope by considering information criteria, theory and logic.

Since the introduction of FIMIX-PLS, a range of alternative latent class techniques have emerged to address some of the limitations of the approach relating, for example, to the technique’s inability to handle heterogeneity in the measurement models and its distributional assumptions. The second part of this article (Part II) discusses alternative latent class techniques in greater detail and calls for the joint use of FIMIX-PLS and PLS prediction-oriented segmentation.

This paper is the first to offer researchers who have not been exposed to the method an introduction to FIMIX-PLS. Based on a state-of-the-art review of the technique in Part I, Part II follows up by offering a step-by-step tutorial on how to use FIMIX-PLS in SmartPLS 3.

This paper aims to investigate a probabilistic mixture model for the nonrigid point set registration problem in the computer vision tasks. The equations to estimate the mixture model parameters and the constraint items are derived simultaneously in the proposed strategy.

The problem of point set registration is expressed as Laplace mixture model (LMM) instead of Gaussian mixture model. Three constraint items, namely, distance, the transformation and the correspondence, are introduced to improve the accuracy. The expectation-maximization (EM) algorithm is used to optimize the objection function and the transformation matrix and correspondence matrix are given concurrently.

Although amounts of the researchers study the nonrigid registration problem, the LMM is not considered for most of them. The nonrigid registration problem is considered in the LMM with the constraint items in this paper. Three experiments are performed to verify the effectiveness and robustness and demonstrate the validity.

The novel method to solve the nonrigid point set registration problem in the presence of the constraint items with EM algorithm is put forward in this work.