Recent studies that analyze scanner panel data often use hierarchical Bayes modeling with dynamic structures and random effects to model consumers' heterogeneity. In this study, we propose a hybrid version of a hierarchical Bayes model with dynamic structures in which both latent classes and random effects are assumed. The proposed model explains consumer heterogeneity as it relates to brand-switching behavior by using latent classes and random effects. This makes it possible to estimate brand-switching behavior accurately by explaining within-class heterogeneity in coefficients with random effects. The proposed method is then applied to an Information Resources Inc. marketing data set with noteworthy results.
This study explains neighborhood differences in people's attitudes toward education by utilizing nationally representative data from Japan. While previous studies have shown that individuals' socioeconomic backgrounds are related to their attitudes toward education, no study in Japan has addressed whether neighborhood differences in socioeconomic characteristics influence these attitudes. Therefore, this study aims to clarify whether a neighborhood socioeconomic factor (percentage of college graduates) differentiates people's attitudes toward education by employing multilevel structural equation modeling techniques. Results indicate some neighborhood differences in individuals' attitudes toward education; the percentage of college graduates in each neighborhood is associated with between-neighborhood differences. In other words, individuals living in a neighborhood with a higher percentage of college graduates have more positive attitudes toward education.
Functional data refer to data that are assumed to be generated from an underlying smooth function varying over a continuum such as time or space. Functional linear models (FLMs) and functional extended redundancy analysis (FERA) are major regression analysis methods for investigating directional associations between predictor and dependent variables that can be functional. In practice, functional data may often arise from heterogeneous subgroups of the population, which involve distinctive directional relationships between predictor and dependent variables. When such cluster-level heterogeneity is present, ignoring this heterogeneity would likely lead to biased statistical inferences. FLMs have been extended to capture cluster-level heterogeneity. Conversely, there has been no attempt to take into account cluster-level heterogeneity in FERA. In this paper, we propose to extend FERA to accommodate cluster-level heterogeneity by combining the method with fuzzy clusterwise regression (FCR) into a unified framework. The proposed method, called fuzzy clusterwise functional extended redundancy analysis (FCFERA), aims to estimate fuzzy memberships of individuals and clusterwise regression coefficient functions at the same time. A penalized least squares criterion is minimized to estimate these parameters by adopting an alternating least squares algorithm in combination with basis function expansions. We conduct simulation studies to investigate the performance of the proposed method. We also apply this method to real data to demonstrate its empirical usefulness.
Since its inception, partial least squares path modeling has suffered from the absence of a single optimization criterion for estimating component weights. A new estimation procedure is proposed to address this enduring issue. The proposed procedure aims to minimize a single least squares criterion for estimating component weights under both Mode A and Mode B. An alternating least squares algorithm is developed to minimize the criterion. This procedure provides quite similar or identical solutions to those obtained from existing Lohmöller's algorithm in real and simulated data analyses. The proposed procedure can serve as an alternative to the existing one in that it is well-grounded in theory as well as performs comparably in practice.