Role of Forgetting in Memory-Based Choice Decisions: A Structural Model

In this paper we develop a general class of dynamic brand choice models, called Lightning Bolt (LB) models, which are consistent with the theory of random utility maximization of consumer choice behavior. The underlying random utility process is Markov, and the inter-temporal evolution of the (utility-maximizing) brand choice process is also Markov. The models permit parsimonious parameterizations of the random utility process in brand choices with the resulting switching probabilities being functionally related to explanatory variables. The model allows for structural state dependence (feedback), habit persistence (inertia), and unobserved heterogeneity. The theoretical development shows that several well known stochastic brand choice models can be deduced from random utility maximization theory. From a managerial perspective, the usefulness of the proposed model stems from its ability to separate out the effects of habits, state dependence and heterogeneity. Strong state dependence effects imply incentives for inducing trial of a brand (e.g., product sampling). In contrast with state dependence, a strong habit persistence effect may be indicative of buyer behavior where inducement of trial of a different brand may not be sufficient to maintain a (sustained) defection of the consumer from the habitually purchased brand to the trial brand. Failure to distinguish between these two effects has important implications. For example, a model that only accounts for state dependence effects would, in the presence of only habit persistence, incorrectly attribute it to state dependence. Based on this the manager could decide to embark on an expensive sampling program that might prove ineffective due to the absence of state dependence. It is also important to distinguish between the effects of unobserved heterogeneity and state dependence. In the absence of true state dependence, failing to account for unobservable variations across households (such as differences in price sensitivities), results in the temporally persistent unobservable elements showing up as state dependence in the model. Hence, a manager may incorrectly opt for sampling as the appropriate marketing action, whereas, a couponing or price promotion strategy should have been preferred. We estimate the model parameters using the AC Nielsen household scanner panel data set on catsup purchases. Further, we investigate empirical techniques for overcoming the “initial conditions” problem that affects many dynamic models of brand choice. Through simulation analysis using the estimated parameters, we show that the calculated profitability of a promotion must take into account the multi-period impact due to state dependence. Further, we demonstrate how ignoring heterogeneity can result in spurious state dependence, thereby making marketing tactics such as product sampling appear far more attractive than they actually are. Specifically, the model that accounts only for habit persistence and state dependence effects predicts a share increase of 2.4% points through sampling for one of the brands in the empirical analysis. Once the effects of unobserved heterogeneity are accounted for, however, this number drops precipitously to 0.3%. While even this low number might fulfill the objectives of the brand manager, it ensures that expectations are not eight times that number. It is important to note that this paper is a first attempt at analyzing the three fundamental dimensions of dynamic brand choice behavior. Our formulation represents a “reduced form” approach as opposed to a “structural” approach. Other limitations of the model include our focus on time-invariant choice sets. Extensions of the model to situations where choice sets vary over time are possible, but are difficult. Another limitation is the dependence on multivariate extreme value distributions. An alternative would be to use multivariate normal distributions, i.e., a probit-like structure which has certain useful properties. A comparison of the two approaches would be useful. Empirical extensions of the model to allow for higher-order processes and nested logit type model structures could also be fruitful. Explicitly incorporating variety-seeking behavior in the model would further enrich the theoretical framework proposed in this paper.

PurposeThe purpose of this study is to investigate the structural relationships between brand consciousness, perceived quality, social influences, traits of vanity, the need for uniqueness (i.e. antecedents), Generation Y purchase intentions and behaviour (consequences) towards luxury fashion goods.Design/methodology/approachAn integrative theoretical model is proposed based on social comparison theory, social impact theory, the perceived quality model and theory of uniqueness to predict the antecedents and consequences of Generation Y luxury fashion goods purchase decisions. Using cross-sectional data, a total of 384 sets of valid questionnaires were collected to perform the statistical analysis for the measurement and structural model using the partial least squares path modelling, a variance-based structural equation modelling technique.FindingsOverall, the structural results imply that the proposed model explains 73.1 and 64 per cent of variances to predict the Generation Y luxury fashion goods purchase decisions. As the several indices for evaluation of goodness of model fit, standardised Root Mean Square Residual, geodesic discrepancy, and unweighted least squares discrepancy show a satisfactory result. The results of two-tailed hypotheses reveal that brand consciousness, perceived quality, social influences, traits of vanity and the need for uniqueness influence Generation Y purchase intention. Moreover, perceived quality and social influences impact purchase behaviour but brand consciousness, traits of vanity and the need for uniqueness do not seem to be significant in explaining the variance in Generation Y purchase behaviour. Furthermore, Generation Y purchase intention is statistically related to purchase behaviour.Originality/valueThere is a lack of empirical evidence and understanding on the influences of consumer purchase intention and behaviour towards luxury fashion goods among the Generation Y. Generation Y is likely to purchase and consume luxury fashion products, and it is important to have a deeper understanding of this market segment.

The disposal of plastic packaging has caused the amount of waste generation, which have a negative impact on the environment. Green packaging, which is more eco-friendly packaging, is harmless to human health and biodegradable. Many businesses have turned to the movement of going green by providing green products and services. Hence, the objective of this study is to investigate the factors that influence the purchase behaviour of consumers towards green packaged products. A total of 200 respondents was recruited using convenience sampling. Data were collected with a set of self-administered questionnaires that covered respondents' profiles and all the variables in the research model. Partial Least Square Structural Equation Modeling was used to assess the measurement model and structural model. Both convergent validity and discriminant validity were deemed satisfactory. The findings have revealed 0.444 of R2 means that the model explains 44.4% of the total variation of purchase behaviour. Environmental concern and environmental attitude were found to affect consumers’ purchase behaviour positively towards green packaged products. Environmental knowledge and social influence, however, were found to be non-significant. This study will be able to provide valuable information to the marketers to focus not only on their products but the packaging of their products as well. This study also provides an insight into the consumers’ purchase behaviour towards green packaged products. It is recommended for future research to recruit more respondents with a more diversified background.

PurposeThis research study seeks to identify what graduating students and alumni perceive to be of most value in courses, and in turn the relationship of those perceptions with the information in evaluations conducted by students at the conclusion of courses.Design/methodology/approachThe project involves empirical research utilizing standard student course evaluation data, and rigorous matching alumni survey data. A focus group, as well as prior academic research, informs the design.FindingsThere are several key conclusions from this study comparing student and alumni perceptions of course satisfaction. Consistent with end‐of‐program survey and focus group observations, career relevance clearly grows with time in importance for determining course satisfaction. Career relevance is not a statistically significant factor for course satisfaction using end‐of‐course student survey responses, but grows to a statistically significant determinant utilizing alumni survey data, larger than both the extent of learning and instructor performance; moreover, instructor performance appears to become less important.Research limitations/implicationsWhile survey responses for individuals as both students and alumni cannot be linked in this study, the high response rate of alumni and the pooling of data suggest results are robust.Practical implicationsIf instructors want students, once they become alumni, to be satisfied with their course experience, they need to teach material which will be truly useful in their careers, even if students do not fully appreciate it during class. And if university presidents want satisfied alumni, they need to ensure their school's curriculum includes material that may be applied in the real world, and that the measures of teaching effectiveness utilized for compensation purposes do not stress too highly traditional measures of student satisfaction.Social implicationsEducators can provide students with a more long‐term satisfying educational experience by ensuring curriculum includes practical material that is truly relevant for careers.Originality/valueStudent evaluations have been commonly used in determining the success of a course, and the effectiveness of their instructors. However, there has only been limited analysis of student evaluations as a measure of what matters most – the benefit to the student once they graduate and move into the working world. Empirical results based on student and alumni survey data identify differences in perceptions between students and alumni, and suggest key recommendations for both instructors and university administrators.

By explicitly accounting for observed and unobserved taste heterogeneity, we investigate state dependence in food consumption on household level. Positive state dependence, which can be interpreted as a loyalty measure, implies habit persistence, whereas negative state dependence implies variety-seeking behavior. Using ConsumerScan data for 2,717 households in Germany for a four-year period (2000 until 2003), we apply mixed multinomial (random coefficient) logit models to analyze consumption behavior in the breakfast cereal market. We find substantial heterogeneity between households: the majority of them express some degree of positive state dependence, which depreciates over time. Merely a small share of the households can be characterized as variety-seekers. The analysis sheds light on the correlation between price sensitivity and loyalty. More state dependent households seem to be more sensitive to price changes. By obtaining household-specific coefficients, we are able to define each household´s position on the distribution of the parameters. Comparing state dependence coefficients and two further loyalty measures (brand-runs and repurchase probabilities), we observe a positive correlation between these, which underpins the importance of the effect of prior consumption. As a diagnostic check, the unconditional distributions of the parameters and the average of the distributions conditional on previous choices are compared.

The study investigates the determinants of green purchase behaviour amongst the millennial cohort in an emerging market of South Africa. The research was theoretically grounded on the diffusion of innovation and the theory of planned behaviour models. A descriptive quantitative research design was followed and primary data was gathered using a self-administered questionnaire. A total of 355 complete responses were suitable for analysis. A partial least squares structural equation modelling technique was used to determine the significance of the interrelationships between the items measuring the study's constructs. Furthermore, the measurement and structural models were assessed. The findings concluded that all the significant independent variables (relative advantage, compatibility and observability) explained 84.0% of the variance in attitude towards green purchase behaviour. Attitudes, subjective norm and perceived behavioural control explained 83.4% of the variance in behavioural intention towards green purchase behaviour and behavioural intention explained 24.2% of the variance in adoption behaviour. It is therefore important for the green industry of South Africa to cultivate the adoption behaviours of green products towards their customer base, thereby securing the further growth of this niche market segment and enhancing purchase behaviours in the long run.

Probabilistic versus random-utility models of state dependence: an empirical comparison

This paper examines demand elasticities using an integrated framework proposed by Hanemann [Hanemann, M. W. 1984. Discrete/continuous models of consumer demand. Econometrica 52(3) 541–561], which models the incidence, brand choice, and quantity decisions of a consumer as an outcome of her utility maximization subject to budget constraints. Although the Hanemann framework has been the mainstay of earlier efforts to examine these decisions jointly, empirical researchers who have used the it to study purchase behavior have often found that the quantity elasticities are around −1, regardless of the brand or category. We attempt to uncover the underlying reasons for this finding and propose approaches to get as close to the “true” quantity elasticities as possible. We do this by (i) analytically demonstrating how assumptions on the distribution of the brand-specific econometrician's errors imply certain restrictions that in turn force quantity elasticities to −1, (ii) discussing how these restrictions can be alleviated by considering a suitable specification of unobserved parameter heterogeneity, and (iii) using scanner data to empirically illustrate the impact of the restrictions on quantity elasticities and the relative efficacy of multiple specifications of unobserved heterogeneity in easing those restrictions. We find that the specification of unobserved heterogeneity crucially influences estimates of quantity elasticities and that the mixture normal specification outperforms the alternatives.

A substantial body of research evidence has now accumulated in the reference price literature. One stream of research has identified the antecedents of reference price and has assessed their effects through experimentation. Others have calibrated a variety of reference price models on panel data and reported the effects on brand choice and other purchase decisions. In this article, the authors review the published literature on reference price in both the behavioral and modeling streams. They offer an integrative framework to review prior research on (1) the formation of reference price, (2) retrieval and use of reference price, and (3) influences of reference price on various purchase decisions and evaluations. In doing so, the authors examine the influences of consumers' prior purchase history and contextual moderators, such as specific purchase occasions, promotional environment of the store, and product category characteristics. They summarize the key findings, identify the unresolved issues, and offer an agenda for further research, which includes a set of testable propositions. They also identify the methodological challenges that face reference price research. In the concluding section, the authors discuss the managerial implications of reference price research.

In this study, we confirmed how social comparison experiences of SNS users affect purchasing behavior and empirically analyzed the mediating effect of beauty interest. Based on domestic and international prior research, we reviewed theories on social comparison experience, beauty interest, and purchasing behavior, and derived a research model and hypothesis based on this. Through online survey data collection, 209 copies were finally analyzed. The results of this study showed that upward comparison of social comparison experiences of SNS users had a significant impact on beauty interest. Beauty interest was found to have a significant impact on purchasing behavior, and downward comparison of social comparison experience was found to have a significant impact on purchasing behavior. And in the verification of the mediation effect, interest in beauty was found to be partially mediated by SNS social comparison experience and purchasing behavior. The types and acceptance of information by consumers using SNS are different, especially the degree of acceptance based on social comparison., In order to secure a variety of consumers, we provide objective information and produce various contents through beauty-related SNS, promote communication and exchange with consumers, provide various types of information, and provide advertising and marketing that can attract the attention of consumers with different self-esteem. It has practical implications in that it is helpful in activities.

E-commerce in Indonesia is currently growing as consumers make an increasing number of online purchases, making the competition fiercer between e-commerce business players in the country. Despite this growth, e-commerce in Indonesia faces challenges remaining to their development such as a gap in internet penetration between rural and urban areas. To win the competition among all e-commerce businesses and face the existing challenges, it is necessary to determine the best strategy to retain current consumers along with getting more and more consumers. Thus, this study aims to better understand the behavior of e-commerce consumers in Indonesia by developing a structural model of purchase intention and behavior based on an adaptation of the unified theory of acceptance and use of technology 2 (UTAUT2) model, with input from other relevant theories and interviews conducted in the country. Data from questionnaires completed by 400 e-commerce consumers are analyzed using covariance-based structural equational modeling (CB-SEM). The CB-SEM analysis finds that purchase intention is significantly influenced by these six factors: facilitating conditions, perceived website quality, security, convenience, economic reasons, and social influence. Furthermore, purchase intention significantly influences the purchase behavior of consumers. Conversely, three factors considered in the study are found not to significantly influence purchase intention: hedonic motivation, variety, and delivery factors. These results help e-commerce businesses to consider important factors when determining key factors influencing consumer purchase intentions and behavior.

Purchase, scanner, and media panels are collections of individuals or households who agree to provide detailed information on their behavior over a period of time. Participants in these panels provide marketers with detailed information on purchase behavior or media use. Panels are designed to represent the demographic and ethnic population of the nation, and range in size from a few thousand to tens of thousands of participants. In purchase panels, participants record information about what they buy. Typically, participants, upon returning from a shopping trip, enter their purchase information into either a diary or an online data collection system. Scanner panels are a special type of purchase panels, where participants are enrolled in a program that automatically captures data on their purchases at retail stores. Scanner panels overcome the limitation of requiring participants to input information manually. Information is necessary on media consumption behavior for the advertisers to successfully place ads where they will be seen by target audiences. That information comes from media panels that consist of households who agree to have their media use tracked by electronic devices. Keywords: purchase panels; scanner panels; media panels; secondary data; advertising; sales promotion

Understanding factors affecting social commerce purchase behavior: A longitudinal perspective

The common theme running through my dissertation work is differentiated products in oligopoly markets with emphasis on the demand analysis. In the markets, consumers generally present heterogeneous preferences even for the same product, which brings additional dimension of differentiation for the market and add more complexity for the equilibrium. The purpose of this dissertation is to qualify and quantify consumers' purchase behavior in relevant markets, and to further investigate the effects on firms' pricing strategy, the market structure, and the effectiveness of policy regulations. The second chapter studies consumers' heterogeneous demands for two transportation fuels, E10 and E85, and evaluates the Renewable Fuel Standard (RFS) program that promotes the consumption of E85. The RFS sets mandate volumes for the consumption of biofuels. Because E85 contains more ethanol than E10, the Renewable Fuel Standard Program conveys a subsidy to E85, relative to E10. A central question for RFS policy concerns the pass-through of the induced subsidy to consumers. To investigate the issue, we develop a structural model suitable to study the pass-through rate at retail level. We consider two dimension of heterogeneity: consumers' randomly distributed locations, which make fuel products provided by different gas stations horizontally differentiated; and their heterogeneous propensities for the two different fuel products, which are vertically differentiated. Accordingly, we build the model from the "Hotelling" framework and extend it to also include the vertical dimension of heterogeneity. The model provides a natural way to represent the role of imperfect competition at the retail level. We then calibrate the model parameters using real world data and simulate the Nash Equilibrium results of the model. Our results show that the pass-through rate is about 0.7 at the baseline level of model parameters. More market power, higher subsidy level, and consumers' higher preferences for subsidized product are related to lower pass-through rate. The third chapter empirically models heterogeneity in the seed demand and also focuses on a crucial issue in farmers' purchase behavior—brand inertia. Brand inertia has long been a topic of interest in the Economics and Marketing literature, and it refers to the situation that an individual is more likely to choose the brand he/she has purchased previously. Among the potential explanations for this tendency, researchers have been particularly interested in the importance of state dependence, defined as the causal dependency of an individual's future choices on their current state. In this chapter, we develop and estimate a micro-level

Labor supply responses to tax changes are a core issue in public economics, reflected by numerous estimates from different methodological approaches. This dissertation contributes to the literature along two main avenues. First, I demonstrate how ex-ante simulations from a static discrete choice model can be validated ex-post by panel data methods estimating the elasticity of taxable earnings (Chapter 2). Second, I analyze the dynamics of labor supply responses to tax changes over time both with regards to panel data approaches (Chapter 3) and incorporated in a structural labor supply framework (Chapter 4). I use Norwegian register and survey data provided by Statistics Norway and focus on labor supply responses in working hours and wage earnings. In Chapter 2, I demonstrate how the standard elasticity of taxable income (ETI) methodology can be used to validate predictions from a structural labor supply model, by analyzing the Norwegian tax reform in 2006. To facilitate comparison, I use the discrete choice model to simulate the effects of the 2006 tax reform on hours of work, and use predicted income levels to obtain a comparable estimate for income elasticity with respect to the net-of-tax rate. In Chapter 3, I analyze the dynamics of earnings responses to tax changes by exploiting substantial exogenous variation in the two-tier surtax schedule for labor income over a period of 14 years (1995-2008). I adopt the dynamic panel data framework by Holmlund and Söderström (2011), and compare with estimates from the conventional static panel approach by Gruber and Saez (2002). I find that the long run responses to tax rate changes in labor supply are about twice as large as the short run responses due to a strong autoregressive effect in earnings. Supplementary conducted simulations suggest that up to 40 percent of the revenue loss due to tax rate cuts in the surtax schedule can be self- financed in the long run by inducing additional generated income. In Chapter 4, I examine the effect of state dependence to policy change responses over time within the framework of an intertemporal structural discrete choice model for women in couples. I allow for state dependence both in preferences and in the set of job opportunities. This causes sluggish labor supply responses to various changes in the wage rate and the tax schedule. Within the first year of a permanent policy change the simulation results suggest that about 50 percent of the full labor supply response can be expected, whereas 90 percent of the response is reached after about 4 years.