ࡱ>  JdJ%` bjbjNN G,,l D"D"D"6-x*3*3*3:4V7'<>=@AAAokpTp,fhhhhhh$chRUD"pjoppAA(~~~p$A:D"A~pf~~f $D"A2= p]1*3sq60'qlD"pp~pppppczppp'ppppD>$b$>b"T<)2- T  Copyright by Son K. Lam, 2009 ALL RIGHTS RESERVED Customer-Brand Identification as a SUSTAINABLE competitive advantage: A Multinational and Longitudinal Examination A Dissertation Presented to the Faculty of the C.T. Bauer College of Business University of Houston In Partial Fulfillment Of the Requirements for the Degree Doctor of Philosophy by Son K. Lam June, 2009 Customer-Brand Identification as a SUSTAINABLE competitive advantage: A Multinational and Longitudinal Examination APPROVED: __________________________________________ Michael Ahearne, Professor of Marketing Chairperson of Committee __________________________________________ Edward Blair, Professor of Marketing __________________________________________ Ye Hu, Assistant Professor of Marketing __________________________________________ Robert Keller, Professor of Management ________________________________________________ Arthur D. Warga, Dean C. T. Bauer College of Business ACKNOWLEDGEMENTS My sincerest gratitude to my dissertation committee chairman, Mike Ahearne. My special thanks to the committee members, Ed Blair, Ye Hu, and Robert Keller, and to the Ph.D. Coordinator, Singin Jimmy. The generous support by Niels Schillewaert is hereby acknowledged. To my family who has always loved me and provided me with tremendous support. To my dad. Customer-Brand Identification as a SUSTAINABLE competitive advantage: A Multinational and Longitudinal Examination Abstract Previous marketing research has been trying to identify a stronger and more enduring predictor of brand loyalty than customer satisfaction in competitive markets. Perceived value, the difference between benefits and costs, appears to be the perfect candidate. Does it? Drawing from the literature on customer-company identification and brand health, this two-essay dissertation proposes that customer-brand identification (CBI), defined as the customers perception, emotional significance, and value of sharing the same self-definitional attributes with a brand, constitutes a sustainable competitive advantage. Compared with perceived value, CBI is expected to be more predictive and enduring in explaining brand health, both current health under normal market conditions and brand resistance under abnormal circumstances such as competitive attacks. Essay 1 examines the relative importance of CBI vis--vis perceived value as the economic driver in predicting customer repurchase intention and customer forgiveness in a cross-sectional setting. Hierarchical linear modeling using a data set of 6,000 consumers from 15 countries shows that cross-sectionally, perceived value is a stronger driver of customer loyalty intention while CBI is more predictive of customer forgiveness. Furthermore, these relationships are generally non-linear, with increasing returns. Surprisingly, national culture interacts more with CBI than with perceived value in predicting customer behavior. Essay 2 complements the first essay by investigating why it is important to build CBI in a competitive, disruptive market setting using a longitudinal design. Results show that when a market is disrupted by an innovative new entrant, CBI saliency serves as a more enduring predictor of switching behavior, an important indicator of customer behavioral loyalty that underlies both the current well-being of a brand and all measures of brand resistance. TABLE OF CONTENTS Abstract v Table of Contents vi List of Tables ix List of Figures x Prologue 1 Consequences of Customer-Brand Identification and Perceived Value: A Multinational Examination Introduction 7 Customer-Brand Identification 10 Social Identity Theory, Identity Theory and Their Marketing Applications 11 Defining Customer-Brand Identification 13 Conceptual Framework and Hypotheses 14 Curvilinear Effects of CBI on Customer Behaviors 14 Curvilinear Effects of Perceived Value on Customer Behavior 17 Relative Importance of CBI and Perceived Value in Predicting Customer Behavior 19 The Moderating Role of National Culture 20 Method 24 Sample 24 Measures 26 Analytical Strategy 27 Results 29 General Discussion 31 Discussion of Findings and Theoretical Contribution 31 Managerial Implications 34 Limitations and Future Research 37 References 40 Appendix 1.A Construct Measures 50 Appendix 1.B Analytical Notes 52 Customer-Brand Identification as a Sustainable Competitive Advantage: A Longitudinal Examination Introduction 56 Customer-Brand Identification 59 Conceptual Framework and Predictions 62 Cross-sectional Effects 62 Longitudinal Effects 64 Method 66 Sample 66 Measures 67 Model Specification 69 Results 72 General Discussion 73 Discussion of Findings and Theoretical Implications 74 Managerial Implications 75 Limitations and Future Research 76 References 81 Appendix 2.A. General Discrete-Time Hazard Model 86 Appendix 2.B. Construct Measures 88 LIST OF TABLES Table 1.1: Means, Standard Deviations, and Intercorrelation Matrix 47 Table 1.2: Hierarchical Linear Modeling Results 48 Table 2.1: Descriptive Statistics 78 Table 2.2: Results of Discrete-Time Hazard Models 79 LIST OF FIGURES Figure 1.1: Conceptual Framework 49 Figure 2.1: Conceptual Framework 80 PROLOGUE Relationship marketing has been the focus of marketing research for decades (Bagozzi 1975; Sheth and Parvatiyar 1995). In this stream of research, the golden standard of sustaining long-term relationships is customer satisfaction. More recently, researchers have challenged the mantra that satisfaction will always lead to customer loyalty, contending that satisfaction is not enough (Jones and Sasser 1995; Oliver 1999). For example, Reichheld (1996) reported that 65%-85% of customers who defect state that they were satisfied or very satisfied before defection. More recently, the quest for a stronger and more enduring predictor of customer loyalty than customer satisfaction in competitive markets has An extensive review of brand health, customer loyalty, and customer-company identification literatures suggests that this variable might be in the form of customer identification with a brand. The health of a brand has been conceptualized in epidemiological terms as consisting of two related yet distinct components: current well-being and resistance (Bhattacharya and Lodish 2000). Brand current well-being is generally reflected in the current market share, baseline sales (i.e., sales when there is no promotion), and customer-based brand equity (Keller and Lehman 2006) under normal conditions. The bulk of the brand loyalty literature focuses on this dimension of brand health, with repurchase intention as the focal criterion variable. Brand resistance refers to the focal brands vulnerability to abnormal fluctuations in the market, such as competitors aggressive promotional campaign, introduction of a disruptively innovative product, or changes in regulations. This vulnerability manifests itself primarily in the form of switching behavior (Bhattacharya and Lodish 2000, p. 8-10), bringing to light the market segment of spurious loyalty (Day 1969; Jacoby and Chesnut 1978). Brand health, therefore, is a broader concept than brand loyalty. It remains unclear, however, as to what variables can serve as valid, persistent antecedents to brand health in the face of disruptive and sustaining forces that are present and active over many consumption episodes (Moore, Wilkie, and Lutz 2002, p. 35). Research on loyalty further suggests that authentic brand loyalty exists only when there is a deeply held commitment to rebuy or repatronize a preferred product/service consistently in the future, thereby causing repetitive same-brand or same-brand-set purchasing, despite situational influences and marketing efforts having the potential to cause switching behavior (Oliver 1999, p. 34, italics added). In this regard, non-spurious loyalty is analogous to the brand resistance dimension of brand health that goes above and beyond mere repurchase intention. Drawing from the customer-company identification literature (Ahearne, Bhattacharya, and Gruen 2005; Bhattacharya and Sen 2003), I propose that customer-brand identification (CBI), defined as the customers perception, emotional significance, and value of sharing the same self-definitional attributes with a brand, is the missing link in predicting brand health even when perceived value and switching costs are controlled for. The brand management literature has postulated several brand concepts. However, CBI is distinct from its predecessor conceptualizations in that CBI reflects and captures the psychological oneness (Ashforth and Mael 1989) while these constructs do not. The fusion of the brand, the self, and self-schemata makes CBI a sticky prior (Bolton and Reed 2004) that might be more enduring than ephemeral satisfaction. Consequently, CBI should be highly predictive across contexts and social settings of several important customer behaviors: in role behavior such as repurchase intention and extra role behavior that goes above and beyond repurchase intention such as forgiving the brand for its transgressions, current behavior and future intentions, support for the identified brand and resistance to competitive attractions. Customer extra role behavior, defined as behavior that go beyond formal role definitions and responsibilities that are generally expected of a customer and are oriented toward helping the firm (Wuyts 2007), might be extremely important in abnormal market conditions such as brand crisis and competitive attacks. In other words, CBI might constitute a sustainable competitive advantage due to its value, rareness, inimitability, and non-substitutability (Barney 1991; Porter 1985; Reed and DeFillipi 1990). Previous research that is based on the conceptual framework of customer-company identification (Bhattacharya and Sen 2003) has received preliminary empirical support that customer-company identification results in higher product utilization and customer extra role behavior such as positive word of mouth (Ahearne, Bhattacharya, and Gruen 2005; Bagozzi and Dholakia 2006; Brown et al. 2005; Donavan, Janda, and Suh 2006). However, there has been little empirical research examining the phenomenon of customer-company identification and customer-brand identification longitudinally or outside of the U.S. More specifically, it remains unclear as to (1) How important it is in the long run, (2) How important it is relative to perceived value, (3) How stable it is in the long run, (4) How it behaves in a competitive environment, and (5) Whether its importance is universal and generalizeable to countries outside of the U.S. This dissertation adopts a strategic application of social identity theory (Ashforth and Mael 1996; Fiol 1991) and builds upon the conceptual framework by Bhattacharya and colleagues (Ahearne, Bhattacharya, and Gruen 2005; Bhattacharya and Sen 2003) to achieve a deeper understanding of CBI and its correlates in three important areas. First, I compare the validity and the functional form of CBI with those of perceived value in predicting customer loyalty and customer forgiveness. Second, I explore the moderating role of national culture of the relationship between CBI and its consequences by adapting Hofstedes (1980, 2001) cultural dimensions. Third, I examine the longitudinal impact of CBI on behavioral loyalty in a competitive context. More specifically, I study how enduring the effect of CBI on customer loyalty is over time in markets where a new entrant tries to uproot customers identification with incumbent brands. This dissertation intends to make four contributions to the marketing literature. First, the longitudinal examination of CBI and its consequences will extend the current understanding of loyalty processes, brand equity, and brand health from the social identity theory perspective. The empirical analyses will help answer three burning questions in the customer-company identification literature: (1) Is CBI merely a metaphor? (2) Is CBI adding any predictive value compared with the golden standards of perceived value and customer satisfaction? and (3) Is CBI a universal phenomenon? Second, this research will be the first to empirically examine national culture as the boundary condition of the relationship between CBI and customer behavior under normal and abnormal conditions (e.g., hypothetical brand transgressions). Third, the close-up look at the customer-brand relationship in normal and disruptive market situations will provide an in-depth juxtaposition of true loyalty versus spurious loyalty. It also complements the existing macro-level literature on innovation and order of entry (e.g., Aboulnasr, Narasimhan, Blair, and Chandy 2008) with an in-depth, micro-level look at the competitive dynamics of technological evolution from the customers perspective. Finally, the explicit incorporation of competition into the customer-company identification literature by applying rigorous analytical methods provides a fresh, micro-level approach that is consistent with recent calls for the incorporation of competition into customer relationship models (Rust, Lemon, and Zeithaml 2004) and for studying relationship marketing during disruptive change. Ultimately, the comprehensive examination of relationship drivers will help managers build vigorous brands and wisely allocate their brand-building resources in the era of globalization. In the next section, I present two essays. The organization of each essay is as follows. I first review the existing literature on customer-company identification and related streams of research, and outline the motivation for each essay. Then, I propose the research questions along with the conceptual framework and hypotheses. This is followed by empirical analyses and discussion of findings. The background literature of this dissertation, namely social identity theory, identity theory, and customer-company identification, is reviewed in depth in Essay 1 and briefly repeated in Essay 2. (Reference for this section appears at the end of Essay 1) Consequences of Customer-Brand Identification And Perceived Value: A Multinational Examination INTRODUCTION Building a strong, healthy relationship with customers tops the Marketing Science Institute (MSI)s 2006-2008 research priorities. Understanding customer behavior resulting from this relationship in tandem with the customers socio-cultural context continues to top MSIs 2008-2010 hot topics. This is not surprising given the consensus that customer retention has a bigger impact on a firms profitability than does customer acquisition (Blattberg and Deighton 1996; Reichheld 1996). Two major ways firms build this relationship are through brands (Aaker and Joachimthaler 2000; Erdem, Swait, and Valenzuela 2006) and value creation (Agustin and Singh 2005; Dodds, Monroe, and Grewal 1991). As Keller and Lehmann (2006, 740) pointed out, Branding has emerged as a top management priority in the last decade. Meanwhile, in the relationship marketing literature, the inter-relationships among perceived value, satisfaction, brand loyalty, and market share figure predominantly (Sheth, Newman, and Gross 1991). While there is consensus that satisfaction is positively related to loyalty, marketing researchers concur that satisfaction is not enough (Oliver 1999). Similarly, Chandrashekaran et al. (2007) found that deep down satisfaction lies satisfaction strength that is critical in translating stated satisfaction into loyalty. In this vein, researchers suggest that perceived value, defined as customers overall assessment of the utility of a product based on perceptions of what is received and what is given (Zeithaml 1988, p. 14), might represent an economic construct at a higher level of abstraction with broader implications for predicting customer loyalty than customer satisfaction (Bolton and Drew 1991). Three questions come to mind: (1) Has this prediction been consistently supported? (2) How does the role of the economic driver change when a psychological driver is included? and (3) Does this hold true for behavior under abnormal conditions such as brand transgressions? Meager empirical research on perceived value has produced mixed results and largely ignored customer behavior other than repurchase intention and willingness to pay. First, while researchers have found that overall satisfaction has an increasing incremental effect on repurchase (Mittal and Kamakura 2001) and willingness to pay more (Homburg, Koschate, and Hoyer 2005), research on the relationship between perceived value and (re)purchase intention has reported inverted-U (Agustin and Singh 2005) and linear (Dodds, Monroe, and Grewal 1991) functions, with effects ranging from strong and positive (Dodds et al. 1991) to marginal ones (Sirohi, McLaughlin, and Wittink 1998). Second, perceived value as a rational consideration of the costs and benefits of staying in the relationship captures only the economic motivation and ignores socio-psychological impetus for customers to enter, maintain, and promote their relationship with the company. From the perspective of marketing as a combination of utilitarian and symbolic exchanges (Bagozzi 1975), focusing on either one of the two exchange drivers, economic or socio-psychological, might bias empirical results. Finally, similar to interpersonal relationships, customer-brand relationships may go through ups and downs, especially during abnormal conditions such as industry crises (e.g., samonella in peanut butter), product recalls (e.g., Mattels toy recall in 2007), and disruptive innovations (e.g., the introduction of the Apples iPhone). Unfortunately, research on customer behavior during abnormal conditions remains sparse. Furthermore, previous research has not identified which customer-brand relationship drivers, economic or psychological, is more important in inducing these behaviors. All of these limitations warrant more investigation. In this study, I focus on the relative importance of perceived value and customer-brand identification as economic and socio-psychological drivers of customer-brand relationship in predicting customer behavior under normal and (hypothetical) abnormal condition in a multinational setting. Specifically, I build on and extend research on customer-company identification, defined as the degree to which customers perceive themselves and the company as sharing the same defining attributes (Bhattacharya and Sen 2003), to the branding literature. In doing so, I rely on the proposition that customer-company identification is the primary psychological substrate for the kind of deep, committed, and meaningful relationships that marketers are increasingly seeking to build with their customers (Bhattacharya and Sen 2003, 76). Consistent with social identity theory (Tajfel and Turner 1985), I define customer-brand identification (CBI) as the customers perception, emotional significance, and value of sharing the same self-definitional attributes with a brand. For the purpose of this study, I treat CBI as a psychological state rather than a process. Theoretically, CBI as a relationship-based construct should be investigated in conjunction with cultural dimensions because national culture strongly influences how individuals appreciate relational and economic benefits (Hofstede 2001). CBI is distinct from its predecessor brand-related conceptualizations in that CBI reflects and captures the notion of psychological oneness (Ashforth and Mael 1989) such that the brand becomes an integral part of the self. The fusion of the brand and the self makes CBI a sticky prior (Bolton and Reed 2004) that is more enduring and self-enriching than mere economic benefits. Consequently, I propose that relative to perceived value, CBI should be more highly predictive of customer behavior under abnormal conditions such as (hypothetical) brand transgressions. From a strategic viewpoint, CBI might constitute a sustainable competitive advantage due to its value, rareness, inimitability, and non-substitutability (Porter 1985). In light of the above discussion, this study seeks the answers to three key questions: (1) How strong is the predictive validity of CBI relative to that of perceived value in explaining customer behaviors under normal and abnormal conditions? (2) What is the functional form of the relationship between CBI and these behaviors; Is this functional form the same across these outcomes; and How does it differ from that exhibited by perceived value? and (3) What is the nature of the moderating effects of cultural orientations on the relationships between CBI and perceived value and these customer behaviors? The answers to these questions not only enrich the limited understanding of the relative importance of customer-brand relationship drivers but also develop theoretical frameworks that are generalizeable across cultures in the era of globalization (Maheswaran and Shavitt 2000). The research also extends the literature on identification beyond the marketing context. In the next section, I first briefly review the theoretical foundation of the construct CBI. I then present the research framework and empirical results from a data set consisting of some 6,000 consumers across 15 countries. The paper concludes with a discussion of findings, theoretical implications, and directions for future research. customer-brand identification Drawing from social identity theory (Tajfel and Turner 1985), Bhattacharya and Sen (2003) argue that while customers are not formal members of companies, they might have self-definitional needs partly filled by companies they patronize, and thus they can identify with a company. As an extension of this logic, in their relationship with brands, customers might also identify with brands since brands represent the actualization of the otherwise abstract, somewhat impersonal existence of the company. This section briefly reviews the theories behind the construct CBI. Social Identity Theory, Identity Theory and Their Marketing Applications Social identity theory (Tajfel and Turner 1985) posits that individuals may define their self-concepts by their connections with social groups or organizations. Based on social identity theory, management researchers develop the concept of organizational identification (Ashforth and Mael 1989; see Ashforth, Harrison, and Corley 2008 for a complete review), defined as the extent to which organizational members define themselves in terms of oneness with the organization. An identity also provides identifiers with a sense of continuity (Albert and Whetten 1985). Individuals who identify strongly with organizations are more likely to engage in identity-congruent behavior, defined as behavior that is consistent with the norms and values of the identity to affirm the identity and to promote the identity. These might include higher in-role performance, embracing organizational values, and extra-role behavior such as voluntarily helping other organizational members in achieving organizational goals (Ashforth and Mael 1989; Ashforth et al. 2008). Marketing research that is based on this theory demonstrated that members of brand communities engage in rituals to extol their beloved brands and help other brand identifiers (Bagozzi and Dholakia 2006; McAlexander, Schouten and Koenig 2002; Muniz and OGuinn 2001). The focus then is more on the collective self or the public self, i.e., the self that is embedded in a collective (e.g., a brand community) or society as a whole (Triandis 1989). At a micro level, identity theory (Stryker and Serpe 1982) focuses on social roles individuals play in various social settings. For example, a student can also occupy the role of a son or daughter and member of a scholar society at the same time. Identity represents the subjective component of a role; identities are organized hierarchically. Identity theory is more concerned about individual behavior and the private self (Triandis 1989). Marketing research based on identity theory is focused on how individual customers behave in agreement with the most salient identity (i.e., highest in the hierarchy) because it provides the most meaning for the self (Arnett, German, and Hunt 2003; Reed 2002). This stream of research also frames customer-product relationship in light of what is me and what is not me (Kleine, Kleine, and Allen 1995). Although social identity theory and identity theory evolve in two fields, social psychology and sociology respectively, these theories share several similar concepts that have been introduced into the marketing literature (Reed 2002). Furthermore, both theories are closely related to the self-concept literature and both examine the connection between the self and society (Belk 1988; Sirgy 1982). Most relevant to this research are identification and identity-congruent behavior. As a side note, identification is different from commitment in that identification possesses the notion of psychological oneness and self-referencing (Ashforth and Mael 1989) while commitment does not. For a detailed conceptual treatment of this topic, see Ashforth et al. (2008). For empirical evidence, interested readers can refer to Bergami and Bagozzi (2000), and Brown et al. (2005). Furthermore, drawing from the work by Bhattacharya, Rao, and Glynns (1995), I posit that all customers who identify with a brand are likely to be loyal to the brand, but all brand loyal customers need not identify with the brand. Defining Customer-Brand Identification Under the overarching theme of relationship marketing (Sheth et al. 1991), previous research on customer-company relationship has been developing along two major streams. The first stream of research is almost exclusively focused on interpretive consumers account about their relationship with brands (Fournier 1998; McAlexander et al. 2002). Theoretically, this stream builds on the literature on the self and close relationships (Aron 2003; Belk 1988; Sirgy 1982). One of the tenets of this school is that possessions can be viewed as the extended self (Belk 1988; Kleine et al. 1995). In other words, this research stream anthropomorphizes brands as relationship partners and views consumer-brand relationships as mostly affect laden (Thomson, MacInnis, and Park 2005). Taking a cognition-based approach that relies primarily on social identity theory (Tajfel and Turner 1985), the second stream of research proposes that customers identify with companies to satisfy one or more self-definitional needs (Ahearne, Bhattacharya, and Gruen 2005; Bagozzi et al. 2008; Bhattacharya and Sen 2003; Einwiller et al. 2006). Most importantly, this identification is not contingent upon interaction with specific organizational members (Turner 1982), or direct experience with the object of identification (Bhattacharya and Sen 2003; Reed 2002). This study builds primarily upon this second stream to examine customer-brand relationship. As I mentioned above, I defined customer-brand identification (CBI) as a psychological state consisting of three elements: cognition (the perception), affect (the emotional significance), and evaluation (the value) that are tied to sharing the same self-definitional attributes with a brand. This conceptualization is consistent with the original tripartite conceptualization in social identity theory (Tajfel and Turner 1985), and integrates the perspectives in organizational identification research (multidimensional, Ashforth and Mael 1989), social categorization theory (mainly cognitive, Turner 1982), and research on close relationships (cognitive and evaluative, Aron 2003). This conceptualization is also in line with the literature on cognition-affect interaction (Zajonc and Markus 1982). Conceptual Framework and Hypotheses Figure 1.1 depicts the conceptual framework. I first focus on the baseline model which captures the simple effects of CBI and perceived value on customer behavioral intentions at the individual level (Level 1). I then lay out the rationale for the moderating effects of national culture at level 2 on these individual-level simple effects. ----- Insert Figure 1.1 about here ----- Curvilinear Effects of CBI on Customer Behaviors CBI might induce two groups of identity-congruent behaviors: behavior during the normal course of the relationship to maintain the identity and behavior that customers might engage in when the identity is questioned, such as brand crisis. Empirical research based on the conceptual framework of customer-company identification (Bhattacharya and Sen 2003) has reported preliminary support that customer-company identification can lead to both customer in-role behavior such as higher product utilization (Ahearne et al. 2005) and extra-role behavior like positive word of mouth, collecting company-related collectibles, and symbol passing (Ahearne et al. 2005; Bagozzi and Dholakia 2006; Brown et al. 2005; Donavan, Janda, and Suh 2006). These extra-role behaviors might exist under normal condition. In the context of customer-brand relationship, I focus on one important type of customer behavior under abnormal condition, customer forgiveness. I define customer forgiveness as customer propensity to overlook and downplay brand transgressions (Aaker, Fournier, and Brasel 2004; Chung and Beverland 2006; Einwiller et al. 2006). I pay particular attention to this extra-role behavior because it may shed light into the distinction between economic-based and psychological-based attachment to a brand. Previous research has established that identification is an antecedent to commitment (Bergami and Bagozzi 2000). Customers who strongly identify with brands develop a deeply-rooted preference for and strong commitment to the brands. Therefore, when faced with identity-inconsistent information such as negative publicity, brand identifiers are more likely to process the information systematically and tend to refute or counter-argue such information as less diagnostic to maintain cognitive consistency (Ahluwalia, Burnkrant, and Unnava 2000; Jain and Maheswaran 2000). In fact, negative information about the brand identity can be considered an identity threat that needs to be addressed for the benefits of the in-group consisting of brand identifiers (Tajfel and Turner 1985). The high level of internalization of the brand into the self also motivates these customers to make generous attributions when transgressions occur, take the perspective of the brand in explaining the abnormality, and take attacks on the brand personally. Research on organizational identification reviewed above predict a linear relationship between identification and outcomes such that the higher individuals identification with a target, the more salient the identity of that target is to them, thus inducing identity-congruent behavior. However, there are theoretical reasons to believe that the relationship between CBI and its consequences might be non-linear. First, customers with low to moderate identification regarded the brand as somewhat detached from the self. Their relationship with the brand might just be very exploratory in nature. It is not until customers perceive the brand as bearing sufficient overlap with their own self when they start to consider the brand as me (Kleine et al. 1995; Sirgy 1982). Once the product/service is embedded inextricably within some portion of the consumers psyche, as well as his or her lifestyle , the consumable is part and parcel of the consumers self-identity and his or her social identity. That is, the person cannot conceive of him or herself as whole without it (Oliver 1999, 40). Second, research on close interpersonal relationship also suggests that one of the benefits of close relationships is that the relationship will grant a partner access to the other partners resources, a phenomenon called self expansion (Aron 2003). In customer-brand relationships, these resources might be the brands resources, associations, and social networks. These social elements both enrich the relationship between the customers and the brand and embed them significantly (Bhattacharya and Sen 2003). At this critical point, the brand identity becomes a stickier and more salient part of the self that drives customers to more actively engage in identity-congruent behaviors (Ashforth and Mael 1989; Bhattacharya and Sen 2003; Bolton and Reed 2004; Stryker and Serpe 1982). At the minimum, customers will develop incrementally stronger repurchase intention for identity maintenance purposes. Furthermore, as CBI surpasses a threshold, customers information processing grows biased in favor of the brand. Selective attention, selective encoding and retrieval, and selective interpretation of information related to the identified brand escalate (Jain and Maheswaran 2000; Raju, Unnava, and Montgomery 2008). Third, once the interdependence between the self and the brand hits this threshold, the motivation to engage in empathetic behavior to reciprocate the relationship partner should substantially intensify (Aron 2003). Reciprocation might come in various forms: intention to buy the same brand again under normal condition, or forgiving the brand for its mistakes. Hence: H1: The relationship between CBI and (a) repurchase intention and (b) customer forgiveness has an increasing incremental effect. Curvilinear Effects of Perceived Value on Customer Behaviors Perceived value provides customers with a rational, economic reason to continue their relationships with the brands. Furthermore, perceived value elevates customer satisfaction which in turn results in higher intention to repurchase and disseminate positive word of mouth (Zeithaml et al. 1996). While the relationship between CBI and customer behavior is driven by identity congruency effects (Stryker and Serpe 1982), the relationship between perceived value and repurchase intention reflects the utility maximization rule (Zeithaml 1988). Previous research suggests that predicting its functional form is complicated. On one hand, Homburg et al. (2005) demonstrate support for disappointment theory (Loomes and Sugden 1986) which suggests that delight and elation resulting from high level of satisfaction should generate increasing incremental value. Mittal and Kamakura (2001) report the same upward curvilinear effect between satisfaction and repurchase behavior. Given the same level of costs, the functional form of the relationship between perceived value and repurchase intention should parallel this increasing incremental pattern. On the other hand, Agustin and Singh (2005) hypothesize a linear relationship between perceived value and loyalty intention by using need-gratification and dual-factor motivation theories (Herzberg 1966; Wolf 1970). Their key arguments are that (1) individuals monovalent needs can be broadly grouped into lower-order, or hygiene needs (e.g., transactional satisfaction) and higher-order, or motivator needs (e.g., trust), (2) beyond certain point of hygiene fulfillment, increasing fulfillment of high-order (lower-order) needs has increasing (decreasing) incremental effects on goal pursuit such as repurchase intention, and (3) perceived value represents a bivalent need that consistently and monotonically motivates goal pursuit regardless of level of fulfillment (Agustin and Singh 2005, 99-100). However, these authors found that, in the nonbusiness airline travel and retail clothing industry, the relationship between value and repurchase intention followed a concave pattern, while that between trust and repurchase intention evidenced a convex trajectory. These mixed findings might be due to two reasons. First, while Agustin and Singhs (2005) study was the first empirical research to examine the simultaneous effects of multiple loyalty determinants, these authors also conjectured that the inclusion of socio-psychological benefits of relational exchanges may provide additional insights into the value-loyalty relationship and that other product categories should be researched. Second, by definition, perceived value places an emphasis on the loss element: satisfaction at what costs? It then follows that the relationship between perceived value, a calculation of losses vis--vis gains, and repurchase intention might reside in the loss domain, which is upward curvilinear in prospect theory (Kahneman and Tversky 1979). This is consistent with disappointment theory mentioned above. When perceived value is high, customers might engage in behavior above and beyond repurchase intention. The underlying mechanism, however, is not so much to promote the brand identity as is true for the case of CBI but rather, to achieve equity in social exchange (Homburg et al. 2005). Customers who appropriate value from the brand might feel the urge to forgive the brand as a token of reciprocation. This urge might grow stronger as value surpasses a threshold that creates elation rather than mere satisfaction of expectations. Hence: H2: The relationship between perceived value and (a) repurchase intention, and (b) customer forgiveness has an increasing incremental effect. Relative Importance of CBI and Value in Predicting Customer Behaviors The nature of customer behavior such as repurchase intention is relationship maintenance, either to maximize economic returns or to maintain a beneficial relationship. This type of customer behavior, however, reflects what Bagozzi (1975) calls utilitarian exchange more than symbolic exchange. Bagozzi (1975, 36) defines a utilitarian exchange as an interaction whereby goods are given in return for money or other goods and the motivation behind the actions lies in the anticipated use or tangible characteristics commonly associated with the objects in the exchange, and symbolic exchange as the mutual transfer of psychological, social, or other intangible entities between two or more parties. Because the relationship between attitude and behavior should be stronger when there exists a match of the level of specificity (Ajzen and Fishben 1977), perceived value as the economic driver should be more important than the psychological driver in predicting behavior that is related to individually-oriented, economic aspects of the exchange between customers and the brand. For high-order goal pursuits that are social and symbolic rather than economic in nature such as brand forgiveness, CBI as the psychological driver should be more important than the economic driver. In identity theory, Stryker and Serpe (1982) call this shared meaning between the identity and identity-congruent behavior. Furthermore, inasmuch as perceived value does not necessarily lead to higher levels of brand internalization to ignite high level of self-sacrifice, I do not expect perceived value to be strongly related to customer behaviors that call for a deep level of information processing. The catalyst behind the increasing incremental effects of CBI is embeddedness, a primarily socio-psychological boost, while that of perceived value is elation, a primarily economic driver. Since embeddedness and elation reside in two different domains, by the same logic of the matching principle (Ajzen and Fishbein 1977), the change in the rate of change (i.e., the acceleration) of the relationship between perceived value and economic-laden behaviors such as repurchase intention must be faster than that between CBI and the same behaviors. Similarly, for social and psychological-laden behaviors that require psychological oneness between the brand and the self (see Ashforth and Mael 1989), the acceleration with which CBI outruns perceived value should be faster. Thus, I hypothesize: H3: Other things equal, when compared with CBI, (a) perceived value is more strongly related to repurchase intention, (b) perceived value is less strongly related to customer forgiveness. The Moderating Role of National Culture Among various conceptualizations of cultural orientations, Hofstedes five cultural dimensions remain the most widely-accepted perspective (Steenkamp, Hofstede, and Wedel 1999). These dimensions include individualism/collectivism, uncertainty avoidance, power distance, masculinity/femininity, and long-term orientation. Most relevant to the context of this research are two cultural dimensions, individualism and uncertainty avoidance. According to Hofstede (2001), individualism refers to the extent to which a culture influences its members to look after themselves or remain integrated into groups, and uncertainty avoidance refers to the extent to which a culture programs its members to feel threatened in unstructured or unknown situations. I focus on these two cultural dimensions because (1) both CBI and perceived value induce customers to maintain and nurture relationships with brands that are beneficial to them socially or economically, and (2) these two cultural orientations drive individuals to focus more on avoiding uncertainty associated with brands they do not identify with or on assigning perceived value different importance weights. Individualism/Collectivism. In their relationship with their brands, customers embedded in an individualistic culture are more likely to be driven by self-serving than self-sacrificing motivation (Aron 2003). Individualistic individuals place more emphasis on variety seeking (vs. belonging), hedonism (vs. survival) while collectivistic individuals value relationship rather than novelty (Roth 1995). Similarly, among brand identifiers, collectivistic consumers are more likely to be embedded with a very enclosed in-group that have stringent norms and expectations and they are more likely to conform to those group norms (Escalas and Bettman 2005). Therefore, collectivistic customers should assign a great deal of importance to relationship maintenance, hence are less likely to switch to other brands from the brands they identify with. In their relationship with acquaintances, collectivistic cultures are likely to attribute failures and abnormalities to external forces such as fate and luck rather than holding the ones they acquaint with responsible for mistakes (Schutte and Ciarlante 1998). Among brand identifiers, those in collectivistic culture value the brands they identify with as closer relationship partners and they are more likely to make generous attributions when their brands make mistakes. Therefore, they might be more forgiving when things go wrong. This suggests: H4: The relationship between CBI and (a) repurchase intention, (b) customer forgiveness will be weaker when national-cultural individualism is high. Perceived value, a computation of cost/benefits, is of the highest importance for customers with individualistic orientation (Triandis et al. 1993). Previous research has suggested that customers in individualistic cultures have high quality expectations and are less tolerant to poor services (Donthu and Yoo 1998). In other words, they are more likely than collectivistic individuals to maintain relationships that provide economic values. Similarly, these individuals might view transgressions of the brands they identify with as unacceptable and therefore are less forgiving. This is because such transgressions raised question about the economic-laden motivation of their relationship with the brands. Thus, H5: The relationship between perceived value and (a) repurchase intention will be stronger when national-cultural individualism is high. However, the relationship between perceived value and (b) customer forgiveness will be weaker when national-cultural individualism is high. Uncertainty avoidance. Countries that are characterized by high uncertainty-avoidance program its individuals to prefer stability, loyalty, and simplicity in consumption. This national-cultural dimension is synonymous with a strong resistance to change and a high need for clarity. Brand identifiers in countries with high uncertainty avoidance should value the importance of the relationships they have with old brands as these brands have lower perceived risk and information costs (Erdem et al. 2006). Hence, among customers who identify with a brand, individuals in these countries will be more likely to be cautious and thorough in making their brand choice and less likely to go through the hassle of brand experimentation (Broderick 2007; Donthu and Yoo 1998). It follows that these brand identifiers will develop stronger repurchase intention and will be willing to sacrifice a bit more to maintain their relationship with the identified brand. Individuals in countries that are characterized by high uncertainty avoidance orientation tend to pay closer attention to the negative aspects of information. When faced with negative publicity, strong brand identifiers in this culture are more likely to process these pieces of negative information as personally relevant, engage actively in information searching, and elaborate them to verify the true merits of the information at hand (Petty and Cacioppo 1981). More importantly, these brand identifiers elaborate the information with the intention to refute it to maintain self consistency rather than to accept it (Ahluwalia et al. 2000; Raju et al. 2008). As a consequence, among brand identifiers, customers embedded in high uncertainty avoidance countries might actually be more likely to forgive the brands when transgressions occur. Hence, H6: The relationship between CBI and (a) repurchase intention, (b) customer forgiveness will be stronger in cultures with high uncertainty avoidance. I mentioned above that the underlying mechanism between perceived value and repurchase intention is to maximize utility and that between perceived value and customer forgiveness is reciprocation. It is risk and structure that are important to individuals in high uncertainty avoidance culture. Given high levels of perceived value, customers in this culture might question whether there is something wrong. For these customers, lower prices might be an indicator of both higher value and low quality (Dodds et al. 1991). Therefore, given high perceived value, customers embedded in high uncertainty-avoidance culture are more likely than those in low uncertainty-avoidance culture to report low repurchase intention. When brands make mistakes, high uncertainty avoidance customers who perceive the brands as delivering high value will be more likely to process the information as given without the motivation to discount its negativity. This will hinder the motivation to reciprocate the brands for the value they have appropriated, and as a result, these customers will be less likely to forgive the brands if the brands make mistakes. Therefore, I hypothesize: H7: The relationship between perceived value and (a) repurchase intention, (b) customer forgiveness will be weaker when national-cultural uncertainty avoidance is high. MethoD I developed a preliminary questionnaire using sample items from Bhattacharya and Sen (2003) and Bagozzi et al. (2008) to measure customer-brand identification and customer forgiveness. I conducted a pretest using a convenience sample of students in a major university in the U.S. These students were asked to fill out a short survey with these scales and commented on the items. After the items were refined, I conducted another large scale pretest in the U.K. with 232 online panel members. I then further refined the wording of the items. The scales I used in the large-scale survey were first written in English. A professional translation service was then tapped to translate all of the scales into eleven languages (Dutch, French, German, Spanish, Polish, Slovakian, Romanian, Danish, Swedish, Italian, and Turkish). The scales were then back-translated to make sure all items were appropriately worded. Then, I asked native speakers of these languages to take the survey and commented on the wording and the length of the survey. I then polished the final version of the online survey, and conducted the survey by sending links to a large online panel. Sample A large international online research firm agreed to grant me access to their proprietary online panel. The link to the online survey was sent to panel members in 15 countries, including Belgium, Holland, France, U.K., Germany, Spain, Italy, Sweden, Denmark, Switzerland, Slovakia, Turkey, Romania, Poland, and the U.S. For each country, a minimum quota of 250 complete responses was set to ensure that there were enough observations and variation within each country. The survey was active for two months. I received complete responses from 5,919 consumers. The unique feature of the data was that it spanned across Scandinavian, Western, and Eastern European countries that have been under-researched. These countries also have considerable variation in terms of national culture. Overall, 46% of the respondents were female, and the average age was 39 (SD = 12.19). The sample size in each country ranged from 202 to 727. I asked these consumers about their relationships with ten brands in five product categories, namely beer, sportswear, cellular phone, fast food chains, and e-commerce sites. These five categories reflected variation in terms of being symbolic, sensory, and functional (Park, Jaworski, and MacInnis 1986; Roth 1995). To ensure that consumers have had enough time to develop their preference and identification with a number of brands, I chose these products because they were at least in the growing phase of their product life cycle. To ensure that the same brands were present across a number of countries and to control for category effect, I focused on corporate brands only (i.e., the name of the company is also the brand) and reserve specific product-brands for future research. At the beginning of the survey, I screened out consumers who did not know the categories well and who did not know the top two brands in the categories well enough (below 3 on 7-point Likert scale). Consumers who passed this hurdle were then randomized to only one brand that they reported they knew well. Measures I measured CBI using a six-item scale that captures three dimensions of identification. The cognitive dimension consisted of two items adapted from Bergami and Bagozzi (2000). The first item in this scale is a Venn-diagram showing the overlap between consumer identity and the brands identity. The second item is a verbal item that describes the identity overlap in words rather than graphically. I measured the affective and evaluative dimensions by two items each, tapping into the affective attachment between the consumer and the brand and whether the consumer thinks the psychological oneness with the brand is valuable to him/her individually and socially. These items were adapted from Bagozzi and Dholakia (2006). I measured perceived value with four items adapted from Dodds et al. (1991). These items focused on the economic value of the brand. Repurchase intention was measured using three items asking consumers how likely they are to repurchase the brand (Zeithaml et al. 1996). Customer forgiveness was measured using two items that asked consumers whether they would forgive the brand for its transgressions. While there have been updates of cultural dimension scores that are generally consistent with Hofstedes measures, I adopted Hofstedes (2001) individualism and uncertainty avoidance scores since the sample included a number of Eastern European countries that were only available in Hofstedes data. I also included brand trust, socio-demographic variables (age, gender) and product categories as control variables. It should be noted that the inclusion of dummies for product categories also controlled for switching costs and other factors that are category specific. Appendix A reports the final sets of measures, with standardized factor loadings. Analytical Strategy Since the data came from consumers across multiple countries, I first conducted exploratory factor analyses and tests on measurement invariance to make sure that consumers understood the scales in a consistent manner. Before estimating the structural models, I also purified scale scores of response styles for each scale in each country, then saved the residualized scale scores (Baumgartner and Steenkamp 2001). For dependent variables, using residualized scores would essentially remove all between-country variation of the level-2 intercept. Therefore, I used raw scores of these dependent variables while controlling for response styles by including the extreme response style and mean response style indices for the corresponding dependent variable in the level-1 regression. I describe and report the results of these steps in detail in Appendix B. Table 1.1 reports the correlation matrix before and after scale purifications, along with construct psychometric properties. ----- Insert Table 1.1 about here ----- The data set included individual consumers that were nested within the countries I sampled from. I used Hierarchical Linear Modeling (HLM, Raudenbush and Bryk 2002) because HLM enables the estimation of individual-level effect while simultaneously controlling for higher-level effects. I first specified a null model (with no predictors at Level 1, and an intercept only at Level 2) to test whether there was significant variation across countries with respect to the dependent variable. Specifically, using raw scores prior to scale purification, I ran two null models in which the intercept of the Level-1 regression predicting one of the two dependent variables is a sum of an intercept and a between-country random error at Level 2. All null models showed that the between-country variance was significant, suggesting that the use of a two-level model was appropriate for testing the hypotheses. I then added the indices of response styles at Level 1 to control for these factors. Then, I proceeded with adding the other control variables, the product categories as dummies, the focal predictors, and their quadratic terms. I selected the quadratic specification over other forms of transformation because it allowed us to test the incremental effect using data that were purified of response biases. The two-level models were as follows: Level 1 DVij = 0j + 1j(ERS) + 2j(MRP) + 3j(GENDER) + 4j(AGE) + 5j(BEER) + 6j(SPORTSWEAR) + 7j(PHONE) + 8j(FOOD) + 9j(CBI) + 10j(VALUE) + 11j(BTRUST) + 12j(CBI)2 + 13j(VALUE)2 + 14j(TRUST)2 + rij . where i = individuals, j = countries, DV = Dependent variables, CBI = Customer-Brand Identification, VALUE = Perceived value, BTRUST = Brand Trust, ERS = Extreme Response Style specific for the DV, MRP = Mean Response Style specific for the DV, rij ~ N(0, 2). Level 2 0j = 00 + 01(IND) + 02(UAI) + u0j, pj = p0 + upj, p = 1-8, pj = p0 + p1(IND) + p2(UAI) + upj, p = 9-11, pj = p0 + upj, p = 12-14. where the Level-2 random effects upj are assumed to be multivariate normal distributed over countries, each with mean of 0, var(upj) = pp, and cov(upj, up j) = pp . When the estimation results suggested that the between-country variation for a particular coefficient was not significant, I constrained the Level-2 random effect for that coefficient to zero, then reran the model. Random effects existed for slope coefficients that involved cross-level interactions between Level-1 predictors and Level-2 national culture. The intercepts for all dependent variables (0j) and the coefficients for response styles (1j and 2j) were specified as random at Level 2. Following Raudenbush and Bryk (2002) s recommendations, all Level-1 predictors were centered within countries, and Level-2 predictors were centered on their corresponding means. I standardized the scores of CBI and perceived value across countries to facilitate interpretation and comparison of their relative importance. The estimation was maximum likelihood. Results Simple effects. Following the steps above, I ran two models for the two dependent variables and reported these results in Table 1.2. Hypothesis H1a predicted that the relationship between CBI and repurchase intention had an increasing effect. The results showed that the linear term was positive and significant ( = .37, p <.000), but the quadratic term was not significant ( = -.01, n.s.). Therefore, hypothesis H1a was not supported. However, I found that the quadratic terms of CBI were significant in predicting customer forgiveness, evidencing an increasing incremental effect of CBI on customer forgiveness ( = .07, p <.01). This result supported hypotheses H1b. As for the economic driver, the positive quadratic terms supported the hypotheses H2a and H2b that there were increasing incremental effects of perceived value on repurchase intention ( = .04, p <.01) and customer forgiveness ( = .04, p <.01). ----- Insert Table 1.2 about here ---- Relative predictive validity of CBI and perceived value. To test hypotheses H3a and H3b, I ran nested models to predict the two dependent variables in which I added only CBI or perceived value to the co-variates only model, and compared their respective explanatory power. The number of parameters estimated for the two dependent variables was different as the Level-2 variation for the slope coefficients (random effects) differed in the two models. The results showed that the model using perceived value as the predictor of repurchase intention fit the data significantly better than the one using CBI as the predictor. Specifically, the improvement in model fit for the model with perceived value as the predictor was 2 (27) = 882.45, p <.00, while that for CBI was 2 (27) = 690.46, p <.00. This means that other things equal, perceived value in fact was a stronger predictor of repurchase intention, in support of H3a. Regarding customer forgiveness, the fit index of the model with CBI as the predictor (2 (15) = 896.84, p <.00) was more significantly improved than that with perceived value as the predictor 2 (15) = 424.01, p <.00). This finding supports the prediction in H3b, that ceteris paribus, CBI in fact was a stronger predictor than perceived value of customer forgiveness. Cross-level interaction effects. Next I tested the cross-level interaction effects between national culture at level 2 and the simple effects at level 1. Hypotheses H4a and H4b predicted that brand identifiers in individualistic countries would report lower repurchase intention and customer brand forgiveness. I found that national-cultural individualism did reduce the positive effect of CBI on repurchase intention ( = -.05, p <.05), and customer forgiveness ( = -.05, p <.05). Therefore, both H4a and H4b were supported. Hypotheses H5a and H5b proposed that national-cultural individualism moderated the effects of perceived value on repurchase intention and customer forgiveness. None of these interactions was significant, thus hypotheses H5a and H5b were not supported. I did not find support for hypotheses H6a about the moderating effect of national-cultural uncertainty avoidance on the relationship between CBI and repurchase intention. Consistent with hypotheses H6b, I found that national-cultural uncertainty avoidance enhanced the relationship between CBI and customer forgiveness ( = .10, p < .01). Both of the interactions between national uncertainty avoidance and perceived value in predicting repurchase intention and customer forgiveness were not significant. Therefore, H7a and H7b were not supported. GENERAL DISCUSSION This study contributes to the marketing literature by examining the relative importance of economic and psychological drivers of customer-brand relationship in predicting two important customer behavior, repurchase intention and customer forgiveness. I tested the conceptual framework using a large data set from 15 countries across the world. Several interesting findings with important theoretical implications emerged. Discussion of Findings and Theoretical Contribution The empirical results clearly suggest that other things equal, perceived value is more predictive of in-role behavior such as repurchase intention than is CBI. However, compared with perceived value, CBI is a more potent predictor of an important extra-role behavior, customer forgiveness. Given the increased likelihood of disruptions in the business environment, the identification of which relationship driver is more important in driving customer in-role and extra-role behavior is not trivial. CBI might represent a unique element of brand equity (Keller 1993) that goes above and beyond repurchase. This study also reveals several nonlinear relationships between these drivers and customer outcomes. Specifically, in predicting repurchase intention, I found that CBI exhibits no incremental effect while perceived value shows an increasing incremental effect. However, in predicting customer forgiveness, both CBI and perceived value exhibit an increasing incremental influence, but CBI exhibited a stronger curvature. Previous research has not examined the relative strength of these curvilinear effects. Here, I found that perceived value in general has a stronger incremental effect when the outcome is primarily utilitarian in nature such as repurchase intention. However, CBI exerted a much stronger incremental effect when the outcomes are mainly symbolic and/or socio-psychological such as customer forgiveness. Comparing the coefficients of CBI, perceived value, and the control variable, brand trust, it is worth noticing that the effects of perceived value and brand trust on customer behaviors were fairly paralleled. CBI, however, outdid both brand trust and perceived value when it comes to customer forgiveness. Furthermore, these patterns together clearly suggest that CBI is theoretically and empirically distinct from both perceived value and brand trust. The linear effect of CBI on repurchase intention seems to suggest that the functional form between CBI and its outcomes may follow different trajectories, depending on whether the outcome serves the purpose of identity-maintenance or identity-promotion. For repurchase intention, which is clearly of an identity maintenance nature, CBI does not exhibit any threshold effect, but rather, has an additive effect above the economic driver of perceived value. For the other outcome, customer forgiveness, improving CBI produces increasing returns. I believe this functional form and this distinction are important not only for future research in marketing but also in management, since most previous work in the literature on organizational identification has not examined these non-linear effects. In addition, I found that perceived value had an increasing incremental effect on repurchase intention. This finding was consistent with disappointment theory and the threshold effect that have received some support in previous research (Homburg et al. 2005; Mittal and Kamakura 2001). However, this functional form was not in line with the decreasing incremental effect documented in Agustin and Singhs (2005) study on airline and retail consumers. In retrospect, this is very consistent with prospect theory (Kahneman and Tversky 1979). More specifically, in categories with lower competitive pressure such as airline, the relationship between value and loyalty might follow a concave function. This is because the perceived value of a brand relative to that of competitors might be too incremental and not obvious to be meaningful to customers. Furthermore, even if there exists relative economic perceived value, the convenience of flight connection might override customers preference for value. On the contrary, most of the categories in this study are in categories with fairly high competitive intensity. A slight increase in perceived value relative to competitor might be very meaningful in pushing customers from a loss framing to a gain framing, from mere satisfaction to elation, hence a convex functional form is observed. This conjecture deserves further exploration. Surprisingly, I found that national culture interacted more with CBI than with perceived value. More specifically, I found that individualism attenuates the effect of CBI on repurchase intention and brand forgiveness. I also found that UAI enhances the positive relationship between CBI and customer forgiveness. These interaction patterns seem to affirm that the CBI construct is social in nature and that uncertainty avoidance and CBI has some congruence in predicting customer outcomes. This is consistent with previous theorization that identification is a strategy that helps reduce subjective uncertainty about ones self concept (Hogg 2003). This finding also suggests that perceived value might be a universal selling point. However, how enduring perceived value is in predicting customer behavior remains to be explored. Managerial Implications This multinational investigation should help brand managers make better decisions on brand investment because the insights will inform managers for which customer segment which variable is a stronger predictor of which customer behavior and more importantly, whether there exists incremental returns for the effects of CBI and perceived value. First, the results provide additional measures of brand health that managers can rely on. Bhattacharya and Lodish (2000) reported that brand managers tend to be overly occupied with tracking market share and ignore other dimensions of brand health. Drawing from the epidemiology literature, they propose that the health of a brand consists of two dimension: (1) the current well-being dimension refers to a brands attraction to consumers in an environment where all brands are operating under normal conditions, and (2) the resistance dimension refers to consumers attraction to the brand when it is under attack from competition or from other elements in the macro-environment. Their brand resistance indices tap into the notion of vulnerability of the focal brand when it is attacked by competitors promotion. I would add that brand resistance can manifest itself in situations that do not involve competitors aggression. Specifically, customer forgiveness is highly relevant to the brand resistance dimension of brand health, but they do not necessarily involve competitive attacks. As an example, when the industry undergoes abnormalities such as product recalls (e.g., the peanut butter product recall in 2008-2009; PETCOs handling of Menu Foods crisis in 2008), the brands resistance dimension is also put to test. In these incidents, extra-role behavior initiated by brand evangelists will serve as an invaluable buffer for the brand to weather the crises. Second, I showed here that relative to perceived value, CBI is a stronger predictor of customer forgiveness. It follows that brand managers who want to build a healthy brand should not only inform customers of the benefits of their brands but also invest resources in building CBI. The identification literature suggests that managers can do so via a number of ways because identification is essentially a function of the prestige of the social entity, including its social responsibility (i.e., the brand, the company), the uniqueness of the social entity, and self-entity congruity (Bhattacharya and Sen 2003). To this end, I would add two important points. First, managers should be aware that while a consistent brand image strategy is desirable, using the same brand across products may easily lead to brand identity anarchy (Aaker and Joachimsthaler 2000), hence diluting the impact of identification on customer behavior. Second, it is not always the case that the brand identity internally intended by the firm is the same as the brand image in the eyes of the external stakeholders such as customers (Dutton and Dukerich 1991). Therefore, managers who are able to quantify this perceptual discrepancy on a frequent basis will be able to enhance customer-brand relationship. One of the intriguing findings of this study was the non-linear effects of CBI and perceived value in predicting customer behavior. First, this implies that managers should keep pushing the limits as there are increasing returns to investment in both CBI and perceived value. Second, the differences in the linear and incremental effects of CBI and perceived value also have clear managerial implications. For example, if the focus of the brand manager is to build repurchase intention, managers should be aware that investment on perceived value can be more beneficial, although building CBI also positively contributes to this outcome. In industries where managers have to rely on the supply chain intermediaries and therefore running a higher risk of making mistakes, my findings suggest that managers should pay attention to building CBI in addition to providing value. On balance, if I take into account of the overall effects of economic and psychological drivers of the customer behaviors examined here, a clear message for brand managers is to build brand identification. Furthermore, the importance of CBI in inducing several important customer behaviors suggests that maintaining a consistent brand identity is of utmost importance. However, in most mergers and acquisitions that take place increasingly often in the real world, the focus has always been on employees transition and largely ignored customers reaction. Brand managers who invest the time in explaining to their customers about the new brand identity in these instances might be able to prevent their customers from switching. Finally, the moderating effects of cultural orientation follow a fairly distinct pattern that managers can utilize. As I mentioned above, national-cultural individualism attenuates the effect of CBI but not perceived value on repurchase intention. The findings seem to suggest that CBI as a psychological driver of customer-brand relationship is not only socially but also culturally-bound while perceived value as an economic driver is fairly universal across individuals in different countries. Therefore, brand managers who operate in countries that are known to be individualistic may want to prime these customers to be more collectivistic. This may take the form of creating forum where these customers can join and have their voice heard (Bhattacharya et al. 1995). By fostering an open communication channel, firms can make these individualistic customers feel like they actually belong, that they are in fact part of a larger group with a common interest in the brand, or a common identity. Furthermore, the interaction pattern also suggests that brand managers should focus their resources on CBI rather than perceived value in countries whose culture is characterized by high uncertainty avoidance. Doing so will help them survive unexpected brand crisis. Limitations and Future Research To the best of my knowledge, this research represents the first multinational study examining an important marketing phenomenon, customer-brand relationship, from the perspective of social identity theory. This study is not free from limitations that might be useful for future research. First, the cross-sectional nature of the study does not allow us to draw conclusion about causality. For example, it might be possible that customer success in one type of behavior engagement and validation of their behavior by the brand in one time period might boost their CBI in the subsequent period. I was constrained by the costs of tracking the same customers over time across many countries. However, within-country analyses show relationship patterns that are consistent with the results I presented here, suggesting that this is not a major limitation. Second, I found here that CBI plays a more important role than perceived value in predicting customer forgiveness. CBI also plays its part in explaining customer in-role behavior such as repurchase intention. Is it possible that customer behaviors are determined by different identification bases at different stages of their consumption (see, for example, Garbarino and Johnson 1999)? A longitudinal study which investigates whether the relationship between the three dimensions of CBI and customer behavior might change over time and its pattern will be useful. In this vein, while there has been some work on the antecedents to perceived value and customer-company identification, there has been no empirical work on the antecedents to the specific dimensions of customer-company identification or CBI. It will be interesting for future research to examine the antecedents to different dimensions, namely cognition, affect, and evaluation, of CBI. What variables can serve as the boundary conditions for these relationships? Will these relationships change over time? All of these questions are academically and managerially interesting. Third, I adopted Hofstedes cultural scores to capture national culture in this study. While Hofstedes data have been frequently used in the marketing literature (Steenkamp et al. 1999), there have been concerns about the richness of its conceptualization. Future research might examine the role of culture using other conceptualizations of culture and values (Rokeach 1973; Schwartz 1992). Finally, this study examines customer-brand relationship in mass markets only. Previous research suggests that in a business-to-business context, relationship is important but buying decision is more rational. However, there has been some preliminary evidence showing that identification can drive customers to engage in extra-role behavior as well (Ahearne et al. 2005). Therefore, it will be interesting to test the hypotheses presented here in a B2B context. In this vein, researchers may want to take into account of multi-levels of identification: identification with the company, with the brand, with the specific stores. Doing so will help refine the current understanding of one of the oldest yet ever-evolutionary phenomenon in marketing, customer-brand relationship. References Aboulnasr, Khaled, Om Narasimhan, Edward Blair, and Rajesh Chandy (2008), Competitive Response to Radical Product Innovations, Journal of Marketing, 72 (2), 94-110. Ajzen, Icek and Martin Fishbein (1977), Attitude-Behavior Relations: A Theoretical Analysis and Review of Empirical Research, Psychological Bulletin, 84 (5), 888-918. Aaker, David A. and Erin Joachimsthaler (2000), The Brand Relationship Spectrum: The Key to the Brand Architecture Challenge, California Management Review, 42 (4), 8-23. Aaker, Jennifer, Susan Fournier, and S. Adam Brasel (2004),When Good Brands Do Bad, Journal of Consumer Research, 31 (1), 1-16. Agustin, Clara and Jagdip Singh (2005), Curvilinear Effects of Consumer Loyalty Determinants in Relational Exchanges, Journal of Marketing Research, 42 (1), 96-108. Ahearne, Michael, C.B. Bhattacharya, and Thomas Gruen (2005), Antecedents and Consequences of Customer-Company Identification: Expanding the Role of Relationship Marketing, Journal of Applied Psychology, 90 (3), 57485. Ahluwalia, Rohini, Robert E. Burnkrant, and H. Rao Unnava (2000), Consumer Response to Negative Publicity: The Moderating Role of Commitment, Journal of Marketing Research, 37 (2), 203-14. Albert, Stuart and David A. Whetten (1985), Organizational Identity, in Research in Organizational Behavior, Vol. 7, ed. Barry Staw and Larry L. Cummings, Greenwich, CT: JAI Press, 26395. Arnett, Dennis B., Steve D. German, and Shelby D. Hunt (2003), The Identity Salience Model of Relationship Marketing Success: The Case of Nonprofit Marketing, Journal of Marketing, 67 (2), 89-105. Aron, Arthur (2003), Self and Close Relationship, in Handbook of Self and Identity, ed. Mark R. Leary and June Price Tangney, New York: The Guilford Press, 442-61. Ashforth, Blake E., and Fred Mael (1989), Social Identity Theory and the Organization, Academy of Management Review, 14 (1), 2039. , Spencer H. Harrison, and Kevin G. Corley (2008), Identification in Organizations: An Examination of Four Fundamental Questions, Journal of Management, 34 (3), 325-74. Bagozzi, Richard P. (1975), Marketing as Exchange, Journal of Marketing, 39 (4), 32-9. , and Utpal M. Dholakia (2006), Antecedents and Purchase Consequences of Customer Participation in Small Group Brand Communities, International Journal of Research Marketing, 23 (1), 45-61. , Massimo Bergami, Gian Luca Marzocchi, and Gabriele Morandin (2008), Customers are Members of Organizations, too: Assessing Foci of Identification in a Brand Community, Working paper. Barney, Jay (1991), Firm Resources and Sustained Competitive Advantage, Journal of Management, 17 (1), 99-120. Baumgartner, Hans, and Jan-Benedict E. M. Steenkamp (2001), Response Styles in Marketing Research: A Cross-National Investigation, Journal of Marketing Research, 38 (2), 143-56. Belk, Russell W. (1988), Possessions and the Extended Self, Journal of Consumer Research, 15 (2), 139-68. Bergami, Massimo and Richard P. Bagozzi (2000), Self-Categorization, Affective Commitment, and Group Self-Esteem as Distinct Aspects of Social Identity in the Organization, British Journal of Social Psychology, 39 (4), 55577. Bhattacharya, C.B., and Sankar Sen (2003), ConsumerCompany Identification: A Framework for Understanding Consumers Relationships with Companies, Journal of Marketing, 67 (2), 7688. , and Leonard Lodish (2000), Towards a System for Monitoring Brand Health from Store Scanner Data, MSI Working Paper, Report No. 00-111. , Hayagreeva Rao, and Mary Ann Glynn (1995), Understanding the Bond of Identification: An Investigation of Its Correlates among Art Museum Members, Journal of Marketing, 59 (4), 46-57. Blattberg, Robert C. and John Deighton (1996), Manage Marketing by the Customer Equity Test, Harvard Business Review, 74 (4), 136-44. Bolton, Lisa E. and Americus Reed, II (2004), Sticky Priors: The Perseverance of Identity Effects on Judgment, Journal of Marketing Research, 41 (4), 397-410. Bolton, Ruth N. and James H. Drew (1991), A Multistage Model of Customers Assessments of Service Quality and Value, Journal of Consumer Research, 17 (4), 375-84. Broderick, Amanda J. (2007), A Cross-National Study of Individual-Cultural Nomological Network of Consumer Involvement, Psychology and Marketing, 24 (4), 343-74. Brown, Tom J., Thomas E. Barry, Peter A. Dacin, and Richard F. Gunst (2005), Spreading the Word: Investigating Antecedents of Consumers Positive Word-of-Mouth Intentions and Behaviors in a Retailing Context, Journal of the Academy of Marketing Science, 33 (2), 12338. Chandrashekaran, Murali, Kristin Rotte, Stephen S. Tax, and Rajdeep Grewal (2007), Satisfaction Strength and Customer Loyalty, Journal of Marketing Research, 44 (1), 153-63. Chaudhuri, Arjun and Moris B. Holbrook (2001), The Chain of Effects from Brand Trust and Brand Affect to Brand Performance: The Role of Brand Loyalty, Journal of Marketing, 65 (2), 81-93. Chung, Emily, and Michael Beverland (2006), An Exploration of Consumer Forgiveness Following Marketer Transgressions, Advances in Consumer Research, 33, 98-9. Dodds, William B., Kent B. Monroe, and Dhruv Grewal (1991), Effects of Price, Brand, and Store Information on Buyers Product Evaluations, Journal of Marketing Research, 28 (3), 307-19. Donavan, D. Todd, Swinder Janda, and Jaebeom Suh (2006), Environmental Influences in Corporate Brand Identification and Outcomes, Journal of Brand Management, 14 (1/2), 125-36. Donthu, Naveen, and Boonghee Yoo (1998), Cultural Influences on Service Quality Expectations, Journal of Service Research, 1 (1), 178-85. Dutton, Jane E., and Janet M. Dukerich (1991), Keeping an Eye on the Mirror: Image and Identity in Organizational Adaptation, Academy of Management Journal, 34 (3), 517-54. Einwiller, Sabine A., Alexander Fedorikhin, Allison R. Johnson, and Michael A. Kamins (2006), Enough Is Enough! When Identification No Longer Prevents Negative Corporate Associations, Journal of the Academy of Marketing Science, 34 (2), 18594. Erdem, Tlin, Joffre Swait, and Aana Valenzuela (2006), Brands as Signals: A Cross-Country Validation Study, Journal of Marketing, 70 (1), 34-49. Escalas, Jennifer Edson and James R. Bettman (2005), Self-Construal, Reference Groups, and Brand Meaning, Journal of Consumer Research, 32 (3), 378-89. Fiol, Marlene C. (1991), Managing Culture as a Competitive Resource: An Identity-Based View of Sustainable Competitive Advantage, Journal of Management, 17 (1), 191-211. Fornell, Claes and David F. Larcker (1981), Evaluating Structural Equation Models with Unobservable Variables and Measurement Error, Journal of Marketing Research, 18 (1), 3950. Fournier, Susan (1998), Consumers and Their Brands: Developing Relationship Theory in Consumer Research, Journal of Consumer Research, 24 (4), 343-73. Garbarino, Ellen and Mark S. Johnson (1999), The Different Roles of Satisfaction, Trust, and Commitment in Customer Relationships, Journal of Marketing, 63 (2), 70-87. Herzberg, Frederick (1966), Work and the Nature of Man. Cleveland: World Publishing Company. Hofstede, Geert (2001), Cultures Consequences: Comparing Values, Behaviors, Institutions and Organizations Across Nations, 2nd ed., Thousand Oaks, CA: Sage. Hogg, Michael A. (2003), Social Identity, in Handbook of Self and Identity, ed. Mark R. Leary and June Price Tangney, New York: The Guilford Press, 462-79. Homburg, Christian, Nicole Koschate, and Wayne D. Hoyer (2005), Do Satisfied Customers Really Pay More? A Study of the Relationship between Customer Satisfaction and Willingness to Pay, Journal of Marketing, 69 (2), 84-96. Jain, Shailendra Pratap and Durairaj Maheswaran (2000), Motivated Reasoning: A Depth-of-Processing Perspective, Journal of Consumer Research, 26 (4), 358-71. Jones, Thomas O. and W. Earl Sasser, Jr. (1995), Why Satisfied Customers Defect, Harvard Business Review, 73 (6), 88-99. Kahneman, Daniel and Amos Tversky (1979), Prospect theory: An Analysis of Decisions under Risk, Econometrica, 47 (March), 263-91. Keller, Kevin Lanes (1993), Conceptualizing, Measuring, Managing Customer-Based Brand Equity, Journal of Marketing, 57 (1), 1-22. Keller, Kevin Lanes and Donald R. Lehmann (2006), Brands and Branding: Research Findings and Future Priorities, Marketing Science, 25 (6), 740-59. Kleine, Susan Schultz, Robert E. Kleine III, and Chris T. Allen (1995), How Is a Possession Me or Not Me? Characterizing Types and Antecedents of Material Possession Attachment, Journal of Consumer Research, 22 (3), 327-43. Loomes, Graham and Robert Sugden (1986), Disappointment and Dynamic Consistency in Choice Under Uncertainty, Review of Economic Studies, 53 (173), 271-82. Maheswaran, Durairaj, and Sharon Shavitt (2000), Issues and New Directions in Global Consumer Psychology, Journal of Consumer Psychology, 9 (2), 59-66. McAlexander, James H., John W. Schouten, and Harold F. Koenig (2002), Building Brand Community, Journal of Marketing, 66 (1), 38-54. Mittal, Vikas and Wagner A. Kamakura (2001), Satisfaction, Repurchase Intent, and Repurchase Behavior: Investigating the Moderating Effect of Customer Characteristics, Journal of Marketing Research, 38 (1), 131-42. Moore, Elizabeth S., William L. Wilkie, and Richard Lutz (2002), Passing the Torch: Intergenerational Influences as a Source of Brand Equity, Journal of Marketing, 66 (2), 17-37. MSI (2006 and 2008), Marketing Science Institute Research Priorities 2006-2008, 2008-2010, http://HYPERLINK "http://www.msi.org"www.msi.org/ Muniz, Albert M., Jr. and Thomas C. OGuinn (2001), Brand Community, Journal of Consumer Research, 27 (4), 412-32. Oliver, Richard L. (1999), Whence Consumer Loyalty? Journal of Marketing, 63 (Special Issue), 33-44. Park, C. Whan, Bernard J. Jaworski, and Deborah J. MacInnis (1986), Strategic Brand Concept-Image Management, Journal of Marketing, 50 (4), 135-45. Petty, Richard E. and John T. Cacioppo (1981), Attitudes and Persuasion: Classic and Contemporary Approaches, Dubuque, IA: William C. Brown. Porter, Michael E. (1985), Competitive Advantage: Creating and Sustaining Superior Performance. New York: Free Press. Raju, Sekar, H. Rao Unnava, and Nicole Votolato Montgomery (2008), The Effect of Brand Commitment on the Evaluation of Nonpreferred Brands: A Disconfirmation Process, Journal of Consumer Research, 35 (5), 851-63. Raudenbush, Stephen W. and Anthony S. Bryk (2002), Hierarchical Linear Models: Applications and Data Analysis Methods. Thousand Oaks, CA: Sage. Reed, Americus, II (2002), Social Identity as a Useful Perspective for Self-Concept Based Consumer Research, Psychology & Marketing, 19 (3), 235-66. Reed, Richard and Robert J. DeFillipi (1990), Causal Ambiguity, Barriers to Imitation, and Sustainable Competitive Advantage, Academy of Management Review, 15 (1), 88-102. Reichheld, Frederick F. (1996), The Loyalty Effect. Boston, MA: Harvard Business School Press. Rokeach, Milton J. (1973), The Nature of Human Values. New York: The Free Press. Roth, Martin S. (1995), The Effects of Culture and Socioeconomics on the Performance of Global Brand Image Strategies, Journal of Marketing Research, 32 (2), 163-75. Rust, Roland T., Katherine N. Lemon, and Valarie A. Zeithaml (2004), Return on Marketing: Using Customer Equity to Focus Marketing Strategy, Journal of Marketing, 68 (1), 109-27. Schutte, Hellmut and Deanne Ciarlante (1998), Consumer Behavior in Asia. Hampshire, UK: MacMillan. Schwartz, Shalom H. (1992), Universals in the Content and Structure of Values: Theoretical Advances and Empirical Tests in 20 Countries, in Advances in Experimental Social psychology, Vol. 25, ed. Mark Zana, Orlando, FL: Academic Press, 1-65. Sheth, Jagdish N. and Atul Parvatiyar (1995), Relationship Marketing in Consumer Markets: Antecedents and Consequences, Journal of the Academy of Marketing Science, 23 (4), 255-271. ----, Bruce I. Newman, and Barbara L. Gross (1991), Why We Buy What We Buy: A Theory of Consumption Values, Journal of Business Research, 22 (2), 159-70. Sirgy, M. Joseph (1982), Self-Concept in Consumer Behavior: A Critical Review, Journal of Consumer Research, 9 (3), 287-300. Sirohi, Niren, Edward W. McLaughlin, and Dick R. Wittink (1998), A Model of Consumer Perceptions and Store Loyalty Intentions for a Supermarket Retailer, Journal of Retailing, 74 (2), 223-45. Steenkamp, Jan-Benedict E. M. and Hans Baumgartner (1998), Assessing Measurement Invariance in Cross-National Consumer Research, Journal of Consumer Research, 25 (1), 78-90. , Frenkel ter Hofstede, and Michel Wedel (1999), A Cross-National Investigation into the Individual and National Cultural Antecedents of Consumer Innovativeness, Journal of Marketing, 63 (2),55-69. Stryker, Sheldon and Richard T. Serpe (1982), Commitment, Identity Salience, and Role Behavior: Theory and Research Example, in Personality, Roles, and Social Behavior, ed. William Ickles and Eric S. Knowles, New York: Springer-Verlag, 199-216. Tajfel, Henri, and John C. Turner (1985), The Social Identity Theory of Intergroup Behavior, in Psychology of Intergroup Relations, ed. Stephen Worchel and William G. Austin, Chicago, IL: Nelson-Hall, 7-24. Thomson, Matthew, Deborah J. MacInnis, and C. Whan Park (2005), The Ties that Bind: Measuring the Strength of Consumers Emotional Attachments to Brands, Journal of Consumer Psychology, 15 (1), 77-91. Triandis, Harry C. (1989), The Self and Social Behavior in Differing Cultural Contexts, Psychological Review, 96 (3), 506-20. Triandis, Harry C., Christopher McCusker, Hector Betancourt, Sumiko Iwao, Kwok Leung, Jose Miguel Salazar, Bernadette Setiadi, Jai B. P. Sinha, Hubert Touzard, and Zbignew Zaleski (1993), An Etic-Emic Analysis of Individualism and Collectivism, Journal of Cross-Cultural Psychology, 24 (3), 366-83. Turner, John C. (1982), Toward a Cognitive Redefinition of the Social Group, in Social Identity and Intergroup Behavior, ed. Henri Tajfel, Cambridge, England: Cambridge University Press, 15-40. Van de Vijver, Fons and Kwok Leung (1997), Methods and Data Analysis for Cross-Cultural Research, Thousand Oaks, CA: Sage Publications, Inc. Wolf, Martin G. (1970), Need Gratification Theory: A Theoretical Reformation of Job Satisfaction/ Dissatisfaction and Job Motivation, Journal of Applied Psychology, 54 (1), 87-94. Wuyts, Stefan (2007), Extra-role Behavior in Buyer-Supplier Relationships, International Journal of Research in Marketing, 24 (4), 301-311. Zeithaml, Valerie A. (1988), Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence, Journal of Marketing, 52 (2), 2-22. , Leonard L. Berry, and A. Parasuraman (1996), The Behavioral Consequences of Service Quality, Journal of Marketing, 60 (2), 31-46. Zajonc, Robert C. and Hazel Markus (1982), Affective and Cognitive Factors in Preferences, Journal of Consumer Research, 9 (2), 123-31. TABLE 1.1 Means, Standard Deviations, and Intercorrelation Matrix Variables12345 1. CBI.89.48.50.57.48 2. Perceived value.53.83.74.70.42 3. Brand Trust.55.76.65.71.39 4. Repurchase Intention.61.73.73.75.44 5. Customer forgiveness.53.46.43.47.64M2.684.384.373.702.98S.D.1.331.531.721.821.52Cronbach alpha.82.95.79a.88.78aComposite reliability.96.95.84.90.78 Notes. All correlation significant at p < .01 (two-tailed). aCorrelation between the two items. CBI: Customer-Brand Identification. Average Variance Extracted on the diagonal. Means and standard deviations were reported for scale scores before scale purification. Zero-order correlations are in the lower matrix; correlations adjusted for response styles are in the upper matrix. TABLE 1.2 Hierarchical Linear Modeling Results PredictorsDependent variablesRepurchase intentionCustomer ForgivenessIntercept componentsIntercept3.55***2.76***Individualism (IND).11-.01Uncertainty avoidance (UAI).06-.02ControlERS-.03***-.02***MRP-.02***.01***Gender (0 = Female)-.08***.04Age.00.00Beer category-.37***.05Sportswear category-.10-.19***Cellular phone category-.19**-.18***Fast food category.24**.06Main effects and Quadratic termsCBI.37***.36***Perceived value.53***.29***Brand trust.60***.14***CBI2-.01.07***Perceived value2.04***.04***Brand trust2.04***.03***Interaction termsCBI x IND-.05**-.05**CBI x UAI.00.10***Perceived value x IND.01.04Perceived value x UAI-.01-.04Brand trust x IND.04-.01Brand trust x UAI-.03.04Pseudo R-square68%36% Notes * p < .10 ** p <.05 *** p <.01. ERS = Extreme Response Style, MRP = Mid-scale Response Style. CBI = Customer-Brand Identification. Unstandardized coefficients. FIGURE 1.1 Conceptual Framework  SHAPE \* MERGEFORMAT  APPENDIX 1.A Construct Measures Construct measuresStandardized factor loadingsCBI  Cognitive dimension ( see below).64 CBI1. Venn-diagram item.78 CBI2. Verbal item.72 Affective dimension.98 CBI3. When someone praises [brand], it feels like a personal compliment..91 CBI4. I would experience an emotional loss if I had to stop using [brand].86 Evaluative dimension.94 CBI5. I believe others respect me for my association with [brand].90 CBI6. I consider myself a valuable partner of [brand].87Perceived value 1. What I get from [brand] is worth the costs.922. All things considered (price, time, and effort), (brand) is a good buy in the (category).913. Compared with other (category) brands, [brand] is good value for the money.934. When I use [brand], I feel I am getting my moneys worth.90Repurchase Intention 1. I consider (brand) my first choice when buying/using (category).902. I will buy more (brand) products or services in the next few years.833. The next time you are going to buy or use [product category], how likely is it that it will be [brand] again? (1: very unlikely, 10: very likely).88Brand Forgiveness 1. When [brand] makes mistakes, I would always forgive.792. Occasional defects, inferior quality, or service failures dont diminish my trust in [brand].80 Cognitive CBI (adapted from Bagozzi and Dholakia 2006; Bergami and Bagozzi 2000) CBI1 (Venn-diagram item)    CBI2. (Verbal item). To what extent does your own sense of who you are (i.e., your personal identity) overlap with your sense of what [brand] represents (i.e., the [brand]s identity)? Anchored by: -4 = Completely different, 0 = Neither similar nor different, and 4 = Completely similar. Note. All scales are 7-point Likert, except CBI1, CBI2, and Item 3 in the repurchase intention scale. APPENDIX 1.B Analytical Notes Exploratory Factor Analysis Consistent with Vijver and Leung (1997), the preliminary analyses started with an exploratory factor analysis (EFA) to explore across countries: the item characteristics, item-total correlations across the measuring items of each construct across countries, and factor loading pattern of all items across all constructs. All scale items showed a consistent pattern across countries did not cross-load heavily onto other factors. This EFA was also necessary before I corrected scale scores for response styles. I then proceeded with confirmatory factor analyses using multiple group models. Measurement Invariance I used multigroup confirmatory factor analyses to test measurement invariance prior to testing the structural model (Steenkamp and Baumgartner 1998). The fit of the configural invariance model in which all loadings were freely estimated yielded good fit (2(1590) = 5,815.78, CFI = .953, TLI = .939, RMSEA = .021, AIC = 7,225.78). All factor loadings were highly significant across all 15 countries, and most of the standardized factor loadings exceeded .60. The full metric invariance model in which all of the factor loadings were constrained to be invariant across countries also yielded good fit (2(1730) = 6,121.68, CFI = .951, TLI = .942, RMSEA = .021, AIC = 7,251.68). Although the increase in chi-square is significant (2(140) = 305.9, p <.00), the alternative fit indices showed only minimal changes. Steenkamp and Baumgartner (1998) recommended that in comparing model fits, researchers should not rely exclusively on the chi-square difference test because of its sensitivity to sample size, and that other fit indices such as CFI, TLI, AIC should also be used. Using this heuristic, I concluded that cross-national invariance was supported. I averaged scale items for each scale to obtain composite scores for each construct. Discriminant Validity Table 1. 2 shows that the square of the pairwise zero-order correlation between any two constructs is smaller than the average variance extracted by the measuring items of the corresponding constructs, demonstrating discriminant validity (Fornell and Larcker 1981). I also tested the discriminant validity of three constructs, CBI, perceived value, and brand trust. I measured brand trust using two 7-point Likert items (I trust this brand, I rely on this brand) adapted from the brand trust scale developed by Chaudhuri and Holbrook (2001). The comparison of average variance extracted and the square of pairwise correlation suggested that discriminant validity was established. The models in which the correlations between the latent constructs were constrained to unity also yielded significantly worse fits, evidencing discriminant validity (2(1) = 1,893, p <.00 for the perceived value - brand trust pair, 2(1) = 4,499.70, p <.00 for the CBI  brand trust pair). It should be noted that the pairwise correlation between CBI and brand trust is moderate ( = .50 after purification), while that between value and brand trust is fairly high ( = .74 after purification). This is consistent with previous research (Agustin and Singh 2005). Correction for Response Styles After examination of the factor structure, I followed the steps Baumgartner and Steenkamp (2001) delineated to purify scale scores of response styles. More specifically, I first calculated the index of extreme response style (ERS), the tendency to endorse extreme responses regardless of content, as the frequency with which a respondent strongly agreed or strongly disagreed with questionnaire items. The index of mid-scale response style (MRP), the tendency to use the middle scale category regardless of content, was computed as the frequency with which respondents endorsed the middle scale category. As recommended by Baumgartner and Steenkamp (2001), I excluded the items belonging to the scale in question when I assessed scale-specific response style indices to avoid confounding of substantive and stylistic variance. Then, I regressed respondents scores on the substantive scales on the corresponding response styles indices, and saved the residualized scores for each scale score for subsequent analyses. These residualized scale scores were essentially been purified of method variance resulting from extreme response styles. Customer-Brand Identification as a Sustainable Competitive Advantage: A LonGITUDINAL EXAMINATION INTRODUCTION Although customer satisfaction figures predominantly in the marketing literature, there have been concerns that customer satisfaction is not enough to predict customer loyalty (Jones and Sasser 1995; Oliver 1999). For example, in a large scale study across several industries, Reichheld (1996) reported that 65%-85% of customers who defect actually state that they were satisfied or very satisfied before defection. Recent search has suggested that perceived value might be a better substitute. Although perceived value is at a higher level of abstraction than satisfaction as it captures customers overall assessment of the utility of a product based on perceptions of what is received and what is given (Zeithamal 1988, p. 14), empirical findings have been mixed (c.f., Dodds, Monroe, and Grewal 1991; Sirohi, McLaughlin, and Wittink 1998). Furthermore, this construct does not capture other non-economic factors such as psychological benefits that might motivate customers to continue to buy what they buy (Sheth, Newman, and Gross 1991). Meanwhile, the branding literature has already suggested that brands can provide benefits above and beyond functional benefits (Keller and Lehmann 2006). However, we know very little about the enduring ability of a brand to create customer lifetime value vis--vis perceived value. In this vein, Bhattacharya and Lodish (2000) propose that the health of a brand can be conceptualized in epidemiological terms as consisting of two related yet distinct components: current well-being and resistance. Brand current well-being is generally reflected in the current market share, baseline sales (i.e., sales when there is no promotion), and customer-based brand equity (Keller 1993) under normal conditions. The bulk of the brand loyalty literature focuses on this dimension of brand health, with repurchase intention and willingness to pay more as the focal criterion variables (Keller and Lehmann 2006). Brand resistance refers to the focal brands vulnerability to abnormal fluctuations in the market, such as competitors aggressive promotional campaign, introduction of a disruptively innovative product, or changes in regulations. This vulnerability manifests itself primarily in the form of switching behavior (Bhattacharya and Lodish 2000, p. 8-10), bringing to light the market segment of spurious loyalty (Day 1969; Jacoby and Chesnut 1978). It remains unclear, however, as to what variables can serve as valid, persistent antecedents to brand health in the face of disruptive and sustaining forces that are present and active over many consumption episodes (Moore, Wilkie, and Lutz 2002, p. 35). Drawing from the customer-company identification literature (Ahearne, Bhattacharya, and Gruen 2005; Bhattacharya and Sen 2003), I propose that customer-brand identification (CBI), defined as the customers perception, emotional significance, and value of sharing the same self-definitional attributes with a brand, is the missing link in predicting brand resistance even when perceived value and switching costs are controlled for. I propose that unlike its predecessor brand concepts, CBI captures the fusion of the brand, the self, and self-schemata, making it a sticky prior (Bolton and Reed 2004) that might be more enduring than either ephemeral satisfaction or manipulable switching costs. In other words, CBI might constitute a sustainable competitive advantage due to its value, rareness, inimitability, and non-substitutability (Barney 1991; Porter 1985; Reed and DeFillipi 1990). My research departs from previous research in several ways. First, previous research on customer-company identification has been cross-sectional. Hence, the dynamic nature of CBI has not been examined. Second, previous research on customer-company identification has largely ignored the important role of competition. Apparently, the call for the incorporation of competition into customer relationship models (Rust, Lemon, and Zeithaml 2004) has not received its well-deserved attention. This is surprising because research in the non-profit marketing literature suggests that participation in activities other than those organized by the focal organization impairs the identification with the focal organization (Bhattacharya, Rao and Glynn 1995). Furthermore, social identity and identity theories suggest that an individual might identify with multiple targets, and only the identity that is the most salient will form the basis for action (Callero 1985; Stryker and Serpe 1994). The salience of an identity to a customer is a function of two major factors, its subjective importance and the situational relevance. A subjectively important identity is one that is highly central to an individuals global or core sense of self.A situationally relevant identity is one that is socially appropriate to a given context (Ashforth and Johnson 2001, p. 32). In competitive markets, this notion of CBI saliency is highly relevant. Finally, there has been little research on the longitudinal effects of customers perceived value on brand loyalty in a competitive setting where value perception is relative rather than absolute. I believe a longitudinal examination of CBI saliency and relative perceived value rather than CBI and perceived value per se should provide useful insights into customer loyalty processes in competitive markets. In light of the above discussion, the goal of this study is to examine the dynamics of CBI saliency when there is a new entrant and its impact on behavioral loyalty at the individual level. In so doing, I seek the answers to two research questions: (1) How predictive is CBI saliency compared with relative perceived value in predicting switching behavior from incumbent brands to the new entrant? and (2) How does this pattern change over time? The organization of this second essay is as follows. First, I briefly review the background literature, and then present the conceptual framework and the hypotheses. Then I describe the empirical context, the sample, and the empirical results. The paper concludes with a discussion of findings, theoretical and managerial implications, and directions for future research. Customer-Brand Identification According to Bhattacharya and Sen (2003), customers might perceive themselves as sharing the same self-definitional attributes with companies. They further posit that, in addition to the array of typically utilitarian values (e.g., high product value, consistency, convenience) that accrue to consumers from their relationship with a company, they believe customer-company identification function as a higher-order and thus far unarticulated source of company-based value (Bhattacharya and Sen 2003, p. 77). I extend this logic to a more micro-level of research domain, brands. This extension is possible for a number of reasons. First, brands should be more familiar to customers than companies, with the exception of corporate brands. Brands represent the concrete actualization of the otherwise abstract companies. Second, a company such as Procter & Gamble might own a multitude of brands in one category, or operate in several product categories. These brands might represent very different brand personality that appeal to very different customer segment. As I mentioned above, I define customer-brand identification as the customers perception, emotional significance, and value of sharing the same self-definitional attributes with a brand. This multidimensional conceptualization is consistent with social identity theory (Tajfel and Turner 1985) from which the customer-company identification literature draws. I recognized that elsewhere identification has been conceptualized as purely cognitive, especially in early work in the organizational identification literature that examines how organizational members identify with companies they work for (Dutton, Dukerich, and Harquail 1994). However, multidimensional conceptualization of identification has recently been gaining acceptance in applied psychology (Ashforth et al. 2008; Harquail 1998; Henry, Arrow, and Carini 1999; Mael and Ashforth 1992) as well as marketing (Bagozzi et al. 2008) literature. This new development in the identification literature is in line with work on the affective and cognitive bases of attitude and the interaction between cognition and affect (Edwards 1990; Fabrigar and Petty 1999). Previous marketing research has examined the notion of identification from two major perspectives. The first school of thought, which is heavily grounded in identity theory (Stryker and Serpe 1982), is more concerned with how role-identity that is associated with social categories such as occupation, gender, or ethnicity guide individuals thinking, judgment and behavior. In this regard, the focus is more on the private self and individual behavior (Reed 2002). For example, in a series of experiments that induce a variety of identities such as parents, teenagers, businessperson, and environmentalists, Bolton and Reed (2004) found that identity-driven judgment is sticky because it triggers an elaborate self-schema and induces social-referencing. They demonstrated that this top-down thinking effect perseveres in the face of subsequent bottom-up analysis (e.g., feature-based analysis) and subsequent counterfactual reasoning, and that strong identifiers are some what less susceptible to social influence effects from group interactions. The second school of thoughts relies more on social identity theory (Tajfel and Turner 1985), focusing on the social and collective self and collective behavior rather than the private self. Examples include research on brand community (Bagozzi and Dholakia 2006), intentional social action (e.g., we-intention versus I-intention, see Bagozzi 2000). A number of researchers have adopted both perspectives since the two original theories, identity and social identity theories, both fall under the umbrella of self research. The customer-company identification literature (Bhattacharya and Sen 2003; Arnett, German, and Hunt 2003) can be classified into this last group as its conceptualization discusses both the private as well as collective aspects of the self, with behavioral outcome ranging from repurchase intention (in-role and personal behavior) to collective action (extra-role behavior such as defending the brand when it is slandered by an out-group such as non-identifiers). Customer-brand identification differs from other brand concepts in important ways. First, it has now been well-established both conceptually and empirically that identification is not the same as commitment (e.g., Ashforth et al. 2008; Bergami and Bagozzi 2000; Brown et al. 2005). Second, identification reflects a psychological oneness resulting from customers active, selective, and volitional act motivated by the satisfaction of one or more self-definitional needs (Bhattacharya and Sen 2003, p. 77). Identification also differs from trust in that customers can trust several brands but might only identify with one brand. Identification differs from loyalty in that identification will lead to loyalty as one form of identity-congruent behavior, behavior that is consistent with the core values and norms associated with the identified target (e.g., the brand). However, not all customers who repurchase the same brand are brand identifiers. It is this last distinction that motivates this research. CONCEPTUAL FRAMEWORK AND PREDICTIONS Drawing from the conceptual work by Sheth, Newman, and Gross (1991) and theorization by Bhattacharya and Sen (2003), I propose that two key group of variables will predict customer loyalty behavior over time, customer characteristics and customer perceptions of both incumbent brand(s) and new competitor(s). In brand healths terminology, this enduring loyalty in face of market change represents an indicator of brand resistance. Although brand resistance can manifest itself in a variety of consumption situations as I mentioned above, here I focus only on brand resistance of incumbent brands when a disruptively innovative brand entered the market. To keep the scope of the study manageable while attending to the longitudinal effects of the focal constructs, I confine the scope of my investigation to only switching behavior from the incumbent brand to the new brand. The conceptual framework is depicted in Figure 2.1. ----- Insert Figure 2.1 about here ----- Cross-Sectional Effects Relative perceived value and customer switching behavior. The relationship between perceived value, defined as the difference between benefits and costs, and repurchase intention has been found to be partially mediated by customer satisfaction (Patterson and Spreng 1997; Whittaker, Ledden, and Kalafatis 2007). Satisfaction is the response to pleasurable fulfillment of needs, desires, and expectations (Oliver 1993). Previous research has shown that satisfaction in turn results in customer loyalty (Oliver 1993). I further argue that it is not perceived value per se that can drive customer behavior. Firms are increasingly adept at creating value for customers. Therefore, a brand that is successful in showing its customers that it offers a little extra than the value other brands can will be able to keep its customers from switching. In fact, customers set their expectation and subsequent satisfaction based on prior experience (Bolton and Drew 1991; Homburg, Koschate, and Hoyer 2005). Similarly, customers who are familiar with incumbent brands use their existing relationship with those brands as a reference point to evaluate new brands. Only brands that can exceed that reference point on the gain side (Kahneman and Tversky 1979) might be able to cause customers to switch. Therefore, I hypothesize: H1: At the time of the introduction of the new brand, the higher the relative perceived value of the incumbent brand, the lower the probability that the customer will switch to the new brand. Customer-brand identification saliency and switching. Customers with high CBI consider the brand to be part of their self. This assimilation is intertwined with customers deep emotional attachment to the brand such that parting with the brand amounts to great psychological distress (Ashforth and Johnson 2001). However, this emotional switching cost is not the only route through which CBI exerts its impact on loyalty behavior. Customers with high CBI enter the phase of determined self-isolation wherein they have generated the focused desire to rebuy the brand and only that brand and acquired the skills necessary to overcome threats and obstacles to this quest (Oliver 1999, p.37-38). Social identity theory and identity theory refer to these behaviors as identity-congruent behavior (Stryker and Serpe 1982; Tajfel and Turner 1985). In its ultimate form, customers entered the phase of immersed self-identity when they participate in brand communities. In other words, CBI carries with it the notion of personal determination and social support that might not be reflected in either switching costs or satisfaction (Oliver 1999, p. 42). An individual may have multiple identifications with different identity targets. For example, a customer might identify with brands that position close to one another. I mentioned above that the social identity and identity theories inform that it is the salience of the identity that forms the basis of behavior. Customers with high CBI consider the brand to be central to their core of self. Yet situational factors might bolster this identification. When the introduction of a new brand takes place in the market space, the existing identity is under threat (Elsbach and Kramer 1996), and therefore customers who identify with incumbents are likely to reconsider their existing relationship with the incumbent brand, making this identity more salient. Yet at the same time, the brand of the new entrant with a novel identity might also be salient. Identity theory suggests that while multiple identifications might exist within an individual, s/he can only be capable of invoking one identity at a time because identities tend to be cognitively segmented and buffered (Ashforth and Johnson 2001; Marks and MacDermid 1996). Even if identities can be simultaneously salient, it is the relative saliency of an identity that matters most in determining which course of action an individual will take (Stryker and Serpe 1982). This suggests the following hypothesis: H2: At the time of the introduction of the new brand, the more salient the CBI with the incumbent brand, the lower the probability that the customer will switch to the new brand. Longitudinal Effects The impact of perceived value might be subject to deterioration due to two reasons. First, competitive attractions are likely to erode the relative perceived value of incumbent brands. Second, perceived value represents essentially a utilitarian driver of consumer-brand relationship; it is not closely linked to the self. Previous research suggests that transaction-specific satisfaction is highly correlated with positive affect (Oliver, Rust, and Varki 1997) but not hot affects. In the aggregate, the effects of perceived value on customer behavioral loyalty, while cognitive and affective in nature, do not reflect a high level of internalization of brand values into the self. On the contrary, the impact of CBI on consumer loyalty is more likely to be enduring over time. Previous experimental research has demonstrated that role identity (e.g., an environmentalist, an executive) can have an enduring effect on judgment, the so-called stickiness effects (Bolton and Reed 2004). In other words, parting with the identity leads to hot affects like psychological distress (Ashforth and Mael 1989). Changing self-related schema does not easily take place, and if it does, it takes time (Bolton and Reed 2004). On this basis, I further propose that the saliency of CBI exerts an enduring effect on customer loyalty behavior. I mentioned above that identity threats such as the introduction of a new brand sensitize the saliency of the identity of existing brands. Over time, if customers find the identity of the new brand to be more attractive, the salience of their old identity might shift to the background. Alternatively, if customers believe their current identity derived from their relationship with the incumbent is superior to the new brands identity, the salience of the existing identity will in fact shift to the fore and grow stronger under identity threat. Taken together, this suggests the following hypothesis: H3: Over time, the effect of the relative perceived value of the incumbent brand on customer switching to the new brand is less enduring (i.e., decays faster) than that of the saliency of CBI with the incumbent brand. Method Sample I collected the data from a large proprietary European online panel of actual consumers in Spain. The research context was the launch of the iPhone 3G is Spain. The launch of the iPhone was particularly suitable for the research questions for several reasons. First, this launch provided a natural starting point of the risks that all consumers were exposed to. Second, the brand was the iPhone 3G, which improved substantially on its prior version 2G. It should be noted that the iPhone 2G was launched in the U.S. in 2007 and was named the innovation of the year by the Times magazine. Third, the cellular phone market in Spain was highly competitive, and switching costs were high as consumers were locked in long-term contracts. These market characteristics provide the most stringent test of the hypotheses. The initial survey was developed in English and was then translated into Spanish by a professional translation service. Two native Spanish speakers checked and completed the wording of the survey. After that, the survey was revised, back translated, and was finally programmed in Spanish. Links to the online survey were then sent to panel members. There were five waves of the surveys. The first wave was conducted two months before the actual launch of the iPhone in Spain. Each wave was live for approximately two weeks. The screening questions in the first wave were whether the panel member owned a cellular phone and their awareness of the launch of the iPhone. The other four waves were carried out at two-month intervals, with the second wave launched approximately 10 days after the actual launch. Therefore, the five waves were equally discretized. To enhance the response rates, panel members were entered into a raffle to win a prize if they completed all waves of the survey. I received 7,050 responses for the first wave. This pool of initial response served as the sample for the last four waves. The average number of responses for each subsequent wave was 2,000. As I will detail later on, I used discrete hazard models to capture switching behavior. I define an event as switching to the iPhone, which I assumed to follow a two-stage process. First, the customer developed the switching ideation. This ideation then may or may not turn into actual switching behavior. With this goal in mind, I compiled a final data set which includes 679 useable responses. These include those who responded to all five waves of surveys, those who expressed an ideation to switch to the iPhone but had not switched, and those who actually switched to the iPhone during the period observed. Customers who switched to a brand other than the iPhone was not recorded as a competing event because the focus of my study was about switching behavior (ideation and/or actual switching) from existing brands to the iPhone. Furthermore, the exclusive service provider for the iPhone in Spain offered quite a few brands, and the switch from an existing brand to another existing brand generally did not rule out the possibility that customers would switch to the iPhone. Among the 679 responses, there were 356 customers who had switching ideation, and 84 actual iPhone-switchers. The sample was diverse in terms of socio-demographics: 37.4% were female; 84.5 lived in urban area; 60.1% were under the age of 30; 86.2% were employed. 48% of the respondents were married, with 87.8% having a bachelor degree. Measures I measured CBI using six items. The cognitive dimension consisted of two items adapted from Bergami and Bagozzi (2000). The first item in this scale is a Venn-diagram showing the overlap between consumer identity and the brands identity. The second item is a verbal item that describes the identity overlap in words rather than graphically. I measured the affective and evaluative dimensions by two items each, tapping into the affective attachment between the consumer and the brand and whether the consumer thinks the psychological oneness with the brand is valuable to him/her individually and socially. These items were adapted from Bagozzi and Dholakia 2006. I measured perceived value with four items adapted from Dodds et al. (1991). Economic and procedural switching costs were measured with five items each, adapted from Burnham et al. (2003) and Jones et al. (2007). Customers rated CBI and perceived value for both their current brand and the iPhone. I constructed the CBI Saliency index by dividing CBI with the incumbent brand by the sum total of CBI for the incumbent brand and the iPhone, then normed this index to be on a scale of seven. Relative perceived value was operationalized in the same way. These relative indices were similar in nature to relative market share in the MCI model (Nakanishi and Cooper 1982). Control variables. Because this study will be tested in the context of a new brand introduction into existing markets, it is necessary to control for customer characteristics related to adoption of innovation. These include customer innate innovativeness (Bass 1969; Steenkamp et al. 1999), and susceptibility to social influence (Bearden, Netemeyer, and Teel 1989). I also controlled for socio-demographic variables (gender, age). Since cellular phone consumption is also involved with a service provider, I controlled for customer satisfaction with the current service provider. In addition to customer perceived value and CBI, switching costs might be another reason why customers keep buying a brand. I included two types of non-relational switching costs: procedural costs and financial costs. Procedural switching costs have been defined as consisting of economic risk, evaluation, learning, and setup costs. Financial switching costs refer to benefits loss and financial-loss costs (Burnham, Frels, and Mahajan 2003). I did not include relational switching costs because it was embedded as a consequence of CBI. This distinction of relational versus non-relational switching costs is consistent with recent theorization by Bendapudi and Berry (1997) that customers may be motivated to maintain relationships with brands because of either constraints such as time and money or dedication such as expected emotional distress if they switch to another brand. Customers with high non-relational switching costs feel trapped as hostage in the relationship with brands (Jones and Sasser 1995). Previous research has consistently found that economic and procedural switching costs are positively related to intention to stay with a service provider (Burnham, Frels, and Mahajan 2003). Model Specification For this study, since the dependent variable is an event (switch/not switch), I adopted survival analysis as the analytical methodology. I specified the logit link (Singer and Willett 2003). It should also be noted that when the probability of the event is low, the complementary log-log link function produces results that are similar to that by the logit link specification (Singer and Willett 2003). I briefly review the general discrete hazard model in Appendix 2.A. Next I applied this framework to develop a two-stage model that was used in this study. I modeled the survey data with five time periods t = 1, 2, , T ; and individuals i = 1, 2, , n. In this model, an individual first demonstrates switching ideation before he or she switches to iPhone. I further define  EMBED Equation.3  QUOTE   when an individual i first indicates switching ideation during time period si. By definition,  EMBED Equation.3  QUOTE   for all s < si  QUOTE   (an individual can only start to form switching ideation once).  EMBED Equation.3  QUOTE   when individual i switches during time period ti  QUOTE  , and similarly,  EMBED Equation.3  QUOTE   for all t < ti  QUOTE   (an individual can switch only once in the duration of the study). Individual i s utility of the focal brand during time period t is  EMBED Equation.3  QUOTE  , where the error term it  QUOTE   is assumed to follow an i.i.d. logistic distribution. The utility governs the individual s behavior in both ideation to switch and actual switching behavior, where actual switching requires the utility to pass a higher bar  > 0 QUOTE  . That is, the individual probability of ideation to switch to the iPhone, given that s/he has not considered it yet, is (1)  EMBED Equation.3  The individuals probability to actually switch, given that s/he has stated switching ideation but has not switched yet, is (2)  EMBED Equation.3  I further specify the deterministic part of the utility as: (3)  EMBED Equation.3  where Dt represents time dummies, CBIS = Customer-brand identification saliency, RPV = Relative perceived value, SWE = Economic switching costs, SWP = Procedural switching costs, PRSAT = Satisfaction with service provider, INNOV = customer innate innovativeness, SUSINF = Susceptibility to information, SUSPUR = Susceptibility to purchasing norm. The effects of covariates are assumed to be stable over time. I used a random-effects model in which the random effect i ~ N (0, 2)  QUOTE   captured the heterogeneity across individuals. To capture the dynamic effects of CBI saliency and relative perceived value, I adopted an AR(1) setting for the coefficients. More specifically: (4)  EMBED Equation.3  (5)  EMBED Equation.3  This approach provides a parsimonious parametric structure with only two parameters, (1) and  QUOTE  , for CBI saliency or relative perceived value. I placed no restriction on the decay parameter  QUOTE   QUOTE   so that its size and sign would reflect the trend and stability of the trend in the time varying coefficients. If  QUOTE   > 1, then the effect of the predictor grows over time. If  QUOTE   = 1, then the effect of the predictor is stable. Finally, if 0 <  QUOTE   <1 , then the effect of the predictor decays over time. If  QUOTE   <0, the effect of the predictor fluctuates between growing and decaying. The likelihood for individual i is: (6)  EMBED Equation.3  The log-likelihood for individual i is (7)  EMBED Equation.3  where si  QUOTE   is the time period when individual i started to consider switching and ti  QUOTE   is the time period when he or she switched. The overall log-likelihood is: (8)  EMBED Equation.3  Results Measurement model. All constructs were first subject to exploratory factor analysis, then confirmatory factor analysis. All of the measurement items loaded onto their intended factor and showed good psychometric properties. The Cronbach alpha reliability index of all constructs exceeded the conventional thresholds. Table 2.1 reports the descriptive statistics of the constructs, and correlation of the two focal predictors, CBI saliency and relative perceived value, over time. I computed composite scores of the focal constructs to run the estimation. Appendix 2.B reports the scale items and reliability indices of each construct measures for each wave. ----- Insert Table 2.1 about here ---- Estimation results. Table 2.2 presents the results for three models: Model A is the baseline hazard model without predictors, Model B is the covariates-only model, and Model C is the final model. In testing the hypotheses, after running Model B, I added the focal predictors, CBI saliency and relative perceived value, one at a time to test whether they improved the overall model fit. These intermediary steps showed that models that models that included either CBI saliency and relative perceived value were superior to the covariates-only model (AIC = 97.7 and 101.1, respectively). This suggested that these two predictors significantly predicted switching behavior. ----- Insert Table 2.2 about here ----- Hypotheses H1 and H2 predict that at the time of the introduction of the new brand, customers who perceive the incumbent brands as providing higher value or possess a stronger identification with the incumbent brand than with the new brand will be less likely to switch. The result in Table 2.1 confirmed both of these hypotheses. More specifically, relative perceived value greatly reduced the switching probability ( = -.814, p <.00), and so was CBI saliency ( = -.356, p <.00). A two-tailed test showed that for the initial stage, the effect of relative perceived value was marginally stronger than that of CBI saliency ( = .457, s.e. = .269, p < .10). It should be emphasized that the dependent variable in the initial stage was only switching ideation and not actual switching. The decay of these effects was also in the direction predicted in hypothesis H3. I found that the effect of CBI saliency was in fact greater than 1 ( = 1.375, p <.05), while the effect of relative perceive value was smaller than 1 ( = .73, p <.05). This suggests that the effect of CBI saliency on actual switching behavior grew stronger over time, while the effect of perceived value on switching waned as time elapses. The growth factor of CBI saliency was much larger than the decay factor of relative perceived value ( = .643, s.e. = .273, p <.05). GENERAL DISCUSSION This is the first empirical test of the longitudinal effect of CBI and value. Its unique features include (1) taking into account the relative strength of these predictors vis--vis the same measures for the new entrant, (2) testing the conceptual framework in a field setting, with a natural occurrence of a critical event that disrupted the market, and (3) controlling for important alternative explanations by important covariates. These features provide the stringent test of theory. It complements the existing macro-level literature on innovation and order of entry (e.g., Aboulnasr et al. 2008) with an in-depth, micro-level look at the competitive dynamics of technological evolution from the customers perspective. Next, I discuss the findings, its implications for theory and managers, and future research avenues. Discussion of Findings and Theoretical Implications In the satisfaction literature, findings about the longitudinal effect of loyalty predictors have been mixed. For example, in a series of longitudinal experiments, Homburg, Koschate, and Hoyer (2006) showed that as the number of experiences increases over time, the influence of cognition increases whereas the influence of affect decreases provided that consumption experiences are consistent. In addition, Garbarino and Johnson (1999) report that satisfaction is the key loyalty driver of newer customers while commitment plays a more important role for older customers. On the contrary, Johnson, Herrmann, and Huber (2006) reported in a three-time period longitudinal study that loyalty intentions are a function of perceived value (largely cognitive in nature) early in the product life cycle and affective attitudes toward the brand become more important driver later on. They however did not control for switching costs. In this vein, it should be noted that Mittal and Kamakura (2001) have shown that the relationship between customer satisfaction and repurchase intention and actual repurchase do not follow the same trajectories. I attributed these mixed findings to the lack of consideration for competitive effects. Here I differ from prior research in important ways by accounting for the competitive effect, considering both switching ideation and actual switching behavior. I found strong support for all of the hypotheses. Cross-sectionally, CBI saliency and relative perceived value appear to be equally strong in predicting customer loyalty. This is consistent with recent findings reported by Vogel, Evanschitzky, and Ramaseshan (2008) and theorization by Lemon, Rust, and Zeithaml (2001). However, the longitudinal effects of CBI saliency and perceived value suggest that the effects of value might erode over time while CBI saliency stands the test of both time and competitive attacks. This study contributed to the burgeoning literature on customer-company identification, and more broadly, relationship marketing. It highlighted the fact that although the relationship between customers and brands might be complexly driven by economic and psychological factors, in the long run, it is the customers psychological bonding with the brand that helps with brand resistance in the long run rather than economic value. This is not to say that perceived value is not important, as evident by the fact that at least cross-sectionally, perceived value relative to the new entrant might help the brand to keep its customers from switching. Having said that, the findings underscore that CBI can serve as a good predictor of brand health (Bhattacharya and Lodish 2001), both for the brands current health and resistance dimensions. In other words, CBI is indeed a sustainable competitive advantage. Managerial Implications Prior research has always been focused on perceived value as the key relationship driver (see for example, , Evanschitzky, and Ramaseshan 2008). The compelling findings of this study have important implications for both brand and customer-relationship managers. For brand managers, the findings suggest that building brand health requires investment on both value delivery and building CBI. Furthermore, such investment must create a tipping point such that the brand has an edge over the competition. Customers may have a liking for a competitor, but doing a little more extra might tip the customers preference in favor of the firm. In addition, it is customer identification with the brand that keeps the customer around for a longer time even when the brand is under attack. The task of brand managers then is to move spurious loyalty customers, those who are prone to competitive promotion and new products, to a higher level of bonding with the brand through socio-psychological ties. These might be as simple as engaging customers in more corporate-sponsored activities, or allowing customers to take part in co-creation such as higher level of customization (Arnett et al. 2003; Bhattacharya and Sen 2003). For customer-relationship managers, the findings suggest that helping customers to see the difference between the firms offerings and those of competitors might pay off nicely. In terms of perceived value, this strategy is not that far away from differentiation suggested by Porter (1985). In terms of customer identification with the brand, enhancing share of mind such that the brand becomes part of the customers self represents a promising and viable strategy. Here the importance is not so much on brand recall and brand awareness, but rather, on whether the brand attracts customers to identify with it such that the focal brand is always more salient than other competing identities and customers are proud about their association with it. In terms of strategic implication, the fact that perceived value might erode faster than CBI does not mean that managers should ignore the former altogether. Quite on the contrary, investment on value might represent an area to focus on to intensify its short-termed effects to poach incumbent brands. Once customers have switched, managers should adapt their strategies to be more relationship-based to build identification. In addition, the weak effect of switching costs suggests that creating high barriers to switch for customers do not pay off in the long run. In fact, such move might create a boomerang effect such that customers calculative commitment turns into bitterness. Limitations and Future Research The results of this study should be interpreted with its limitations in mind. First, I only focused on one product category. Although this focus has the advantage of controlling for industry-level effects, the generalizability of the results might be compromised. However, given that the study context is fairly stringent (high switching costs, high competition) and the empirical analysis provides very strong support of theory, I believe the results should hold in many other contexts. The enduring effect of CBI saliency is not trivial. However, I have not examined the antecedents to this saliency. The organizational identification literature (Ashforth and Mael 1989) and recent research on customer-company identification (Arnett et al. 2003) seems to suggest that organizational prestige is one of the most important factors that create higher identification. However, higher identification does not guarantee high saliency, and it is shown here that saliency matters. Therefore, I suspect that brand uniqueness may play a more important role than brand prestige when the targets of identification are brands. Broadly, this would suggest that CBI has some peculiarities that are not available to draw from findings in the organizational literature. Much more research is needed to explore this interesting area. In the same vein, although the stickiness of CBI is a contribution to the literature, I have not uncovered the nuances of the underlying processes. For example, which part of CBI is sticky? Is it cognition, affect, or evaluative aspects? Which aspect is more important? Does this pattern change over time? The answer to these questions should prove useful for theoretical advancement. In this vein, if CBI is important, it might be interesting for future research to explore how CBI aspects evolve over time and what managers can do improve specific aspects of CBI. Finally, due to cost concern, I was able to track customers over a year. Future studies that extend the duration might reveal deeper insights into the customer-brand relationship. TABLE 2.1. Descriptive Statistics Means and Standard Deviations VariablesM1 SD1M2 SD2M3 SD3M4 SD4M5 SD5Overall Mean SDCBI Saliency 3.77 .683.80 .653.82 .673.89 .693.90 .713.84 .68Relative perceived value3.82 .683.83 .693.84 .723.89 .703.91 .693.86 .70Switching costs Economic3.55 1.463.56 1.433.58 1.424.12 1.524.01 1.393.67 1.47Switching costs Procedural 3.46 1.293.59 1.283.68 1.334.02 1.323.99 1.293.65 1.30Satisfaction with service provider4.63 1.294.53 1.274.56 1.254.78 1.254.81 1.194.62 1.26Innate Innovativeness*3.79 .77Susceptibility information*4.16 1.31Susceptibility purchasing norms*2.01 1.18 Notes: * Personality trait, measured at time 1 only. CBI = Customer-Brand Identification. Mt = mean for each time period, SDt = standard deviation for each time period, t = 1-5. For each row, the first line is the mean, and the second line shows the standard deviation. Correlation Matrix Variables123456789 1. CBI Saliency t1 2. CBI Saliency t2.54 3. CBI Saliency t3.53.75 4. CBI Saliency t4.46.74.78 5. CBI Saliency t5.48.65.72.76 6. Relative perceived value t1.50.44.40.37.36 7. Relative perceived value t2.27.57.52.52.43.46 8. Relative perceived value t3.30.43.58.51.45.47.68 9. Relative perceived value t4.28.44.52.62.50.45.61.6810. Relative perceived value t5.30.46.48.54.67.42.55.60.61 Notes. All correlations are significant at p <.01 TABLE 2.2 Results of Discrete-Time Hazard Models Parameter EstimatesModel AModel BModel CIntercept-1.267*** (.095)-2.343*** (.463)1.769*** (.786)D2-.075 (.131)-.022 (.132)-.218 (.565)D3-.075*** (.162)-.667*** (.164)- .744 (.805)D4-1.229*** (.194)-1.136*** (.196)- .693 (.919)D5-1.373*** (.207)-1.243*** (.210)-.064 (1.182)Switching costs - Economics-.014 (.042).065 (.044)Switching costs - Procedural-.120** (.050)-.083* (.050)Satisfaction with service provider-.046 (.043).005 (.045)Consumer innate innovativeness.357*** (.060).309*** (.063)Susceptibility - Information-.049 (.045) .027 (.047)Susceptibility Purchasing norms.043 (.050)-.003 (.052)Gender.253** (.117).257** (.123)Age-.002 (.006)-.000 (.006)CBI Saliency-.356*** (.131)CBI Saliency decay CBIS1.375*** (.170)Relative perceived value-.814*** (.186)Relative perceived value decay RPV.732*** (.159)Switching threshold .958*** (.144)1.112*** (.147)1.511*** (.157)Random effect .358*** (.105).305*** (.114).359*** (.109)Goodness-of-fit-2LL2486.42432.12295.9Parameters estimated71519AIC2500.42462.32333.9BIC2525.12515.12400.9Notes. * p <.10; ** p <.05; *** p <.01. Standard errors in parentheses. FIGURE 2.1 Essay 2 - Conceptual Framework  SHAPE \* MERGEFORMAT  References Aboulnasr, Khaled, Om Narasimhan, Edward Blair, and Rajesh Chandy (2008), Competitive Response to Radical Product Innovations, Journal of Marketing, 72 (2), 94-110. Ahearne, Michael, C.B. Bhattacharya, and Thomas Gruen (2005), Antecedents and Consequences of Customer-Company Identification: Expanding the Role of Relationship Marketing, Journal of Applied Psychology, 90 (3), 57485. Arnett, Dennis B., Steve D. German, and Shelby D. Hunt (2003), The Identity Salience Model of Relationship Marketing Success: The Case of Nonprofit Marketing, Journal of Marketing, 67 (2), 89-105. Ashforth, Blake E. and Scott A. Johnson (2001), Which Hat to Wear? The Relative Salience of Multiple Identities in Organizational Contexts, Michael A. Hogg and Deborah. J. Terry (Eds.), Social Identity Processes in Organizational Contexts, Philadelphia: Psychology Press, 31-48. , and Fred Mael (1989), Social Identity Theory and the Organization, Academy of Management Review, 14 (1), 2039. , Spencer H. Harrison, and Kevin G. Corley (2008), Identification in Organizations: An Examination of Four Fundamental Questions, Journal of Management, 34 (3), 325-74. Bagozzi, Richard (2000), On the Concept of Intentional Social Action in Consumer Behavior, 27 (3), 388-396. Bagozzi, Richard and Utpal M. Dholakia (2006), Antecedents and Purchase Consequences of Customer Participation in Small Group Brand Communities, International Journal of Research Marketing, 23 (1), 45-61. , Massimo Bergami, Gian Luca Marzocchi, and Gabriele Morandin (2008), Customers are Members of Organizations, too: Assessing Foci of Identification in a Brand Community, Working paper. Barney, Jay (1991), Firm Resources and Sustained Competitive Advantage, Journal of Management, 17 (1), 99-120. Bass, Frank M. (1969), A New Product Growth for Model Consumer Durables, Management Science, 15 (5), 215-227. Bearden, William O., Richard G. Netemeyer, and Jesse E. Teel (1989), Measurement of Consumer Susceptibility to Interpersonal Influence, Journal of Consumer Research, 15 (4), 473-481. Bendapudi, Neeli, and Leonard L. Berry (1997), Customers Motivations for Maintaining Relationships with Service Providers, Journal of Retailing, 73 (1), 15-37. Bergami, Massimo and Richard P. Bagozzi (2000), Self-Categorization, Affective Commitment, and Group Self-Esteem as Distinct Aspects of Social Identity in the Organization, British Journal of Social Psychology, 39 (4), 55577. Bhattacharya, C.B., and Sankar Sen (2003), ConsumerCompany Identification: A Framework for Understanding Consumers Relationships with Companies, Journal of Marketing, 67 (2), 7688. , and Leonard Lodish (2000), Towards a System for Monitoring Brand Health from Store Scanner Data, MSI Working Paper, Report No. 00-111. , Hayagreeva Rao, and Mary Ann Glynn (1995), Understanding the Bond of Identification: An Investigation of Its Correlates among Art Museum Members, Journal of Marketing, 59 (4), 46-57. Bolton, Lisa E. and Americus Reed, II (2004), Sticky Priors: The Perseverance of Identity Effects on Judgment, Journal of Marketing Research, 41(4), 397-410. Bolton, Ruth N., and James H. Drew (1991), A Multistage Model of Customers Assessments of Service Quality and Value, Journal of Consumer Research, 17 (4), 375-384. Brown, Tom J., Thomas E. Barry, Peter A. Dacin, and Richard F. Gunst (2005), Spreading the Word: Investigating Antecedents of Consumers Positive Word-of-Mouth Intentions and Behaviors in a Retailing Context, Journal of the Academy of Marketing Science, 33 (2), 12338. Burnham, Thomas A., Judy K. Frels, and Vijay Mahajan (2003), Consumer Switching Costs: A Typology, Antecedents, and Consequences, Journal of the Academy of Marketing Science, 31 (2), 109-126. Callero, Peter L. (1985), Role-Identity Salience, Social Psychology Quarterly, 48 (3), 203-215. Day, George S. (1969), A Two-Dimensional Concept of Brand Loyalty, Journal of Advertising Research, 9 (3), 29-35. Dodds, William B., Kent B. Monroe, and Dhruv Grewal (1991), "Effects of Price, Brand, and Store Information on Buyers' Product Evaluations," Journal of Marketing Research, 28 (3), 307-19. Dutton, Jane E., Janet M. Dukerich, and Celia V. Harquail (1994), Organizational Images and Member Identification, Administrative Science Quarterly, 39 (2), 23963. Edwards, Kari (1990), The Interplay of Affect and Cognition in Attitude Formation and Change, Journal of Personality and Social Psychology, 59 (2), 202-216. Elsbach, Kimberly D. and Roderick M. Kramer (1996), Members Responses to Organizational Identity Threats: Encountering and Countering the Business Week Rankings, Administrative Science Quarterly, 41 (3), 442-476. Fabrigar, Leandre R. and Richard E. Petty (1999), The Role of the Affective and Cognitive Bases of Attitudes in Susceptibility to Affectively and Cognitively Based Persuasion, Personality and Social Psychology Bulletin, 25 (3), 363-381. Garbarino, Ellen and Mark S. Johnson (1999), The Different Roles of Satisfaction, Trust, and Commitment in Customer Relationships, Journal of Marketing, 63 (2), 70-87. Harquail, Celia V. (1998), Organizational Identification and the Whole Person: Integrating Affect, Behavior, and Cognition, in Identity in Organizations: Buidling Theories through Conversations, David A. Whetten and Peter C. Godfrey, eds. Thousand Oaks, CA: Sage, 171-207. Henry, Kelly Bouas, Holly Arrow, and Barbara Carini (1999), A Tripartite Model of Group Identification: Theory and Measurement, Small Group Research, 30 (5), 558-581. Homburg, Christian, Nicole Koschate, and Wayne D. Hoyer (2005), Do Satisfied Customers Really Pay More? A Study of the Relationship between Customer Satisfaction and Willingness to Pay, Journal of Marketing, 69 (2), 84-96. Jacoby, Jacob and Robert W. Chestnut (1978), Brand Loyalty. New York: John Wiley & Sons. Johnson, Michael D., Andreas Herrmann, and Frank Huber (2006), The Evolution of Loyalty Intentions, Journal of Marketing, 70 (2), 122-132. Jones, Michael A., Kristy E. Reynolds, David L. Mothersbaugh, and Sharon E. Beatty (2007), The Positive and Negative Effects of Switching Costs on Relational Outcomes, Journal of Service Research, 9 (4), 335-355. Jones, Thomas O. and W. Earl Sasser, Jr. (1995), Why Satisfied Customers Defect, Harvard Business Review, 73 (6), 88-99. Kahneman, Daniel and Amos Tversky (1979), Prospect theory: An Analysis of Decisions under Risk, Econometrica, 47 (March), 263-91. Keller, Kevin Lanes (1993), Conceptualizing, Measuring, Managing Customer-Based Brand Equity, Journal of Marketing, 57 (1), 1-22. Keller, Kevin Lanes and Donald R. Lehmann (2006), Brands and Branding: Research Findings and Future Priorities, Marketing Science, 25 (6), 740-59. Lemon, Katherine N., Roland T. Rust, and Valerie A. Zeithaml (2001), What Drives Customer Equity, Marketing Management, 10 (1), 20-25. Marks, Stephen R. and Shelley M. MacDermid (1996), Multiple Roles and the Self: A Theory of Role Balance, Journal of Marriage & Family, 58 (2), 417-432. Mittal, Vikas and Wagner A. Kamakura (2001), Satisfaction, Repurchase Intent, and Repurchase Behavior: Investigating the Moderating Effect of Customer Characteristics, Journal of Marketing Research, 38 (1), 131-42. Moore, Elizabeth S., William L. Wilkie, and Richard Lutz (2002), Passing the Torch: Intergenerational Influences as a Source of Brand Equity, Journal of Marketing, 66 (2), 17-37. Nakanishi, Masao and Lee G. Cooper (1974), Parameter Estimation for a Multiplicative Competitive Interaction Model-Least Squares Approach, Journal of Marketing Research, 11 (4), 303-311. Oliver, Richard L. (1999), Whence Consumer Loyalty? Journal of Marketing, 63 (4), 33-44. (1993), Cognitive, Affective, and Attribute Bases of the Satisfaction Response, Journal of Consumer Research, 20 (3), 418-430. , Roland T. Rust, and Sajeev Varki (1997), Customer Delight: Foundations, Findings, and Managerial Insight, Journal of Retailing, 73 (3), 31136. Patterson, Paul G. and Richard Spreng (1997), Modelling the Relationship between Perceived Value, Satisfaction and Repurchase Intentions in a Business-to-Business, Services Context: An Empirical Examination, International Journal of Service Industry Management, 8 (5), 414-434. Porter, Michael E. (1985), Competitive Advantage: Creating and Sustaining Superior Performance. New York: Free Press. Reed, Americus, II (2002), Social Identity as a Useful Perspective for Self-Concept Based Consumer Research, Psychology & Marketing, 19 (3), 235-266. Reed, Richard and Robert J. DeFillipi (1990), Causal Ambiguity, Barriers to Imitation, and Sustainable Competitive Advantage, Academy of Management Review, 15 (1), 88-102. Reichheld, Frederick F. (1996), The Loyalty Effect. Boston, MA: Harvard Business School Press. Rust, Roland T., Katherine N. Lemon, and Valarie A. Zeithaml (2004), Return on Marketing: Using Customer Equity to Focus Marketing Strategy, Journal of Marketing, 68 (1), 109-27. Sheth, Jagdish N., Bruce I. Newman, and Barbara L. Gross (1991), Why We Buy What We Buy: A Theory of Consumption Values, Journal of Business Research. 22 (2), 159-170. Singer, Judith D. and John B. Willett (2003), Applied Longitudinal Data Analysis, Oxford, NY: Oxford University Press Inc. Sirohi, Niren, Edward W. McLaughlin, and Dick R. Wittink (1998), A Model of Consumer Perceptions and Store Loyalty Intentions for a Supermarket Retailer, Journal of Retailing, 74 (2), 223-45. Steenkamp, Jan-Benedict E. M., and Katrijn Gielens (2003), Consumer and Market Drivers of the Trial Probability of New Consumer Packaged Goods, Journal of Consumer Research, 30 (3), 368-384. Steenkamp, Jan-Benedict E. M., Frenkel ter Hofstede, and Michel Wedel (1999), A Cross-National Investigation into the Individual and National Cultural Antecedents of Consumer Innovativeness, Journal of Marketing, 63 (2), 5569. Stryker, Sheldon and Richard T. Serpe (1982), Commitment, Identity Salience, and Role Behavior: Theory and Research Example, in Personality, Roles, and Social Behavior, William Ickles and Eric S. Knowles, eds. New York: Springer-Verlag, 199-216. Tajfel, Henri and John C. Turner (1985), The Social Identity Theory of Intergroup Behavior, in Psychology of Intergroup Relations, Stephen Worchel and William G. Austin, eds. Chicago, IL: Nelson-Hall, 7-24. Vogel, Verena, Heiner Evanschitzky, and B. Ramaseshan (2008), Customer Equity Drivers and Future Sales, Journal of Marketing, 72 (6), 98-108. Whittaker, Graham, Lesley Ledden, and Starvros P. Kalafatis (2007), A Re-examination of the Relationship between Value, Satisfaction, and Intention in Business Services, Journal of Services Marketing, 21 (5), 345-357. Zeithaml, Valerie A. (1988), Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence, Journal of Marketing, 52(2),2-22. APPENDIX 2.A General Discrete Time Hazard Model Let the time period i for individual i is the censoring time, at which either the event yit happens or the observation stops. The event yit takes on the value of either 1 or 0. A discrete hazard likelihood can be written as: (A.1)  EMBED Equation.3  where ci =1  QUOTE   if the data is right censored. hit  QUOTE   is the hazard rate for individual i at time period t. In discrete hazard analysis,  QUOTE  hit  is also the probability for discrete hazard. The probability that the event happens at QUOTE   i is simply (A.2)  EMBED Equation.3   QUOTE  The likelihood function is: (A.3)  EMBED Equation.3  and the log-likelihood function is: (A.4)  EMBED Equation.3  (A.5)  EMBED Equation.3  When ci =1  QUOTE  , 1- ci = 0 QUOTE  . Its straightforward to show that (A.6)  EMBED Equation.3  Therefore, the censoring indicator ci  QUOTE   does not appear in the final LL. (A.7)  EMBED Equation.3  Finally, the log likelihood function can be rewritten as: (A.8)  EMBED Equation.3  APPENDIX 2.B Construct Measures CBI with incumbent brands and with Iphone (Bagozzi and Dholakia 2006; Bergami and Bagozzi 2000.  =.58/.85/.87/.87/.88,  =.66/.86/.87/.88/.87 for CBI with the iPhone and the current brand, respectively) Cognitive CBI CBI1. (Venn-diagram item).    CBI2. (Verbal item). To what extent does your own sense of who you are (i.e., your personal identity) overlap with your sense of what [brand] represents (i.e., the [brand]s identity)? Anchored by: -4 = Completely different, 0 = Neither similar nor different, and 4 = Completely similar. Affective CBI (7-point Likert, strongly disagree/strongly agree) CBI3. When someone praises [brand], it feels like a personal compliment. CBI4. I would experience an emotional loss if I had to stop using [brand] Evaluative CBI (7-point Likert, strongly disagree/strongly agree) CBI5. I believe others respect me for my association with [brand] CBI6. I consider myself a valuable partner of [brand]. Perceived Value of the incumbent brand and the iPhone (Dodds, Monroe, and Grewal 1991; 7-point Likert, strongly disagree/strongly agree. &*0EF     N j ~       4 i   9 ǿϛ||xtxt||m h]h;B&h4XkhL%h \hO h]h h]h-5h]hU35h]hP\5 h]hWd h]h_ h]hfl h]hQFh]haV5h]h~5 h]haVh]h~5; h]h@9 h]h@7 h]h~h\p, h]hV|*     12dgd~$a$gd~ H!$gdV|$ H!$a$gdaVn  ! " 9 \ q r s t w         $ H!$a$gdaVdgd~ $da$gd~ ? h  . / 0 1  @ H!$gd_ gd~ gd~    T  d o p $ H!$a$gd0$ H!$a$gd~ H!$dgd;B& d`gdL%dgdO d`gd~$ H!$a$gd- @ H!$gd_9 a d j m p %([ai\fs|+l}0LOWÿѭ}wlh]hr -CJaJ hL%CJh]hmCJ h\p,CJ hc/CJh]h8CJh]hLCJh]h~CJ h]h+ h]h \h \h@7 h]h@7 h]hC6 h]h0h]h~5h]h~5; h]hY h]h~hL%hO( O !dgd!1$ H!$da$gd-d1$`gdL% d`gd~$ H!$a$gd0W_{_ !",O)<bntm깮Ϲꗏ h]h8h]hC65h]hgM5 h]h-h38CJaJh#CJaJh]hj`:CJaJh]hCJaJh]hLCJaJhL%CJaJh]h8CJaJh\p,CJaJh]hr -CJaJh]hmCJaJ1STbefgi 4/01=>LM\]wx!/01㩢 h]h(h( h]h h# h]hMn h]hC6h!1hMn5h!1hC65 h]hgM h]h;I h]h.% h]hx: h]h0= h]hN G h]hh!1hjF h]h W78Tg 4j1?N^y 1Y H!$dgd!1 H!$dgd!1dgd!113VX|}~?@xy;<Z[WYZûҴ h]hV, h]htb h]hL h]hh]h5h]h-5 h]h!1h!1hMn h]hMnh!1h85h!1hMn5h0 h]hC6h(>YAz=\ !dgd( !dgd!1 !dgd!1 H!$dgd!1Z 8!dgdtb !dgd$ !da$gd !dgd(Z`bcef  &LNOPQYZ[gq%&OP{h]h8gmHnHu h]h h]h]# h]hG h]hO h]h8gh]h .5h]hV,5 h]hV,hGmh5hGmh-5 h]h h0h( h]hC6 h]htbh!1 h]hL h]h/Z)QZ &)/2X6 d`gdO d1$`gd d1$`gdr - d`gd8g$ H!$da$gdreb$ 8!da$gdtb 8!dgdtbP\{   ) B R n!o!r!s!!!!0"["c"i""#C#K#$%%%&& &&$&2&'w''''8((q) *+***.+T+,),@,,Ҷ h]h7 h]h h]hr -mHnHu h]hGmhh]hr -6hh]h]#5h'| h]h]# h]h8g h]hr - h]hOhY( h]hG h]hU:,-Q.11112272B2r2233334H4z44445;5<5B5W6X6D7p7r7_8y88A9f999999$:):8:A:J:::T;v<<<<=$=,=-=D=w=====V>>>>>>?;?>????žԾž h]hechzEh h]hz h]h9 h]hOh h]h]#h]hr -6hOc h]hr -KX69=?"@#@$@%@&@'@(@)@*@+@,@-@.@/@0@1@2@3@4@5@$a$gdN H!$dgd~ d`gdec d1$`gd9???!@"@#@(@8@<@@@@@@@@yBBB"EAEF!FFFGG HbHcHHHMNNOO_PPPPQQž񳫤~wpwplh!6 h]hD5 h]h|$d h]hzhfg h]h`?h0-Kh]h?#6hY(h h]h?#h]hC65 h]h#hN h]hNh]hN5:h]h95;h]hN5; h]h h]hr -h[; h]hechzE*5@6@7@8@9@:@;@<@=@>@m@@@@@@@@@@@@@@@D $da$gdC6 d`gd?#gd[;$a$gdNDHOUW[]%]___eh&mMmrv*vwwwwz|$a$gd?# d1$`gd@7dgd?# $da$gdN d1$`gdN d1$`gd?#QSSS7SSSVVVɤ5=>׷)»M $da$gdN d1$`gd! %^`%gd?#1$^`gdz d1$`gd?#dgd?#gd?b ^`gd?#ǫ~5MYfoWYio<=> ˳3׷ )*+m»nžӾkڿ3眣hQy h]h!hfhGmh]h?#6 h]h`E h]hh?# h]h?bh]hTf5h]h?#5 h]hzhz h]h?# h]hh h]hD$hD$;3?MNOu=jl{)u e-./#*Ba!HNpr{ʰʰhj8hS h]hb h]h-& h]h!6h?#hRh# h]h!h]h?#5; h]h?bh]hTf5h]h?#5 h]h?# h]hD5?.r{OnX hdh1$^hgd?# & F dh1$gd?#d7$8$H$`gd?# $d1$a$gd?# d1$gd?# d1$`gd?# d1$`gdNdgd?# d1$`gd!{| !&0FGKLOPZiB]qruUWq~Ilngrknjh]h?#\h]hD5\ h]h:Unh]h?#5hk h]h9hj8 h]hDh]h?#6 h]h?# h]hD5GCb  *.06@DFR\`bhtxz ^bdjtz| $*,68:BFtxh]h?#H*h]h?#H*h]h?#6h]h?#6\h]h?#\h]hD5\ h]h?#Nh|6l Z\*x d1$gdQy $da$gd?#dgd?# d1$`gd?#1$gd?# & F dh1$gd?# d`gd?#gd?# hdh1$^hgd?#x8BjvPTZ`pr  ,068<FJRVnr| NRʾʾh]h?#\h]h?#6mHsHh]h?#mHsHh]h?#H*mHsHh]h?#H*h]h?#H*h]h?#6 h]h?#HR 48>DNT *Z^!@t */3>F~>|0(8PVx)<EPR¾hGmhhQyhQyhQy6hGj9h]h?#6h]h?#5hk h]hD5h]h?#H* h]h?#KR "@BFf$dfjl:<@B^bfptY 4THhƿڻڻڻڻڻڻh h]hD5h]h?#6hh?#6hQy h]h?#hGj9h]h6h]hH* h]hhJRgrs>BNR   FGKbw-9I> e n y { | - L c h   B C \ _    8 E d _m۾hKOh h]hNh) h]h9h]h?#5jhQy0JUhQyh]h?#6 h]hD5 h]h?#h h]hEG| C _w:#&)y..b6:dgd?# d1$`gdN d1$gdN d1$`gd?# d1$gd?# $da$gdN d1$`gdQym1huJuz)P<6I]^ ';> ?  ""]#b#.$M$%&&&(ۼȸȴȴhhGmh(h]hN6 h]hNh]hN5h) h]hD5h]h?#5 h]h9hmIhKO h]h?#C()))a*v*?.x.y......g/l/Y0e00000R1n11111I2]222b6j66677x::::;;;;'<=<<<T=p==Ҽh]h?#6mH sH h]h?#mH sH h]hLmH sH h]hL6 h]hLh]h?#5;\h_ h]h9 h]h?#h]h?#5h h]hD5 h]hNh]hNmHnHu4::;;P<Q<<<==(>)>????@@OAPAA07$8$H$^`0gd?# 0^`0gd?##0d1$^`0gd?#01$7$8$H$^`0gdL$d7$8$H$a$gd_===>>>???@5@@@)A=APAAAPBlBCCaCuC D8DNDDDEESEhEzE{EFFnFFLGpGH,Hд~t~m h]h<h]h<6aJh]h<aJh]h?#B*phh]h?#6 h]h?#h]h?#0J$h]h?#0J$6h]h?#]mHnHuh]h?#6mHnHuh]h?#mHnHuh]h?#6mH sH h]h?#mH sH h]h?#0J+mH sH *AA}B~B,CCCIDJDE EzE{E/F0FFFGG=H01$^`0gd?#"$0d^`0a$gd?#%0d00XDYD^`0gd< 0^`0gd?#07$8$H$^`0gd?#=H>HHHIIJJJJ\KLLMMMMNN#O$O"$0d^`0a$gd?# 0^`0gd?# 0d [$\$^`0gd?#01$^`0gd?#,HHHfIzIIJJJJJ.KJKKKLLMMMM\NpNNO"O#O$OOOePPPQQQQRR9SMSSSS}TTT,UIUJUUUzVVVVWxW{W}WWWX YYYh]h?#H*h]h?#]h]h?#6]h]h]#6 h]h]# h]h h?#h]h?#0J+6h]h?#0J+hYh]h?#6 h]h?#A$OOOPP!Q"QQQRR^S_SSTZU[UUUVV0^`0gdL01$^`0gd?# 0^`0gd?#"$0d^`0a$gd?#07$8$H$^`0gd?#VVVWW;X*B*phjh:h?#Ujh:h?#U h:h?#h:h?#6h]hL6] h]hLh]h?#0J+6h]h?#0J+h]h<6 h]h<h]h?#6 h]h h?# h]h?#0\\`]a]]]^^]_^_````aaaabb 0^`0gd?#h1$7$8$H$^h`gdL"$0d1$^`0a$gd?#01$7$8$H$^`0gd?#"$0d^`0a$gd?#bJbbbbbbcc ccc+dmddddde eeeeeefJfdffg*g9ggggh'hhh9iiii_j{jjjkkTlpl*m>mNmOmh]h<6 h]h< h]h h]hKOh]h<6mHnHuh]h<mHnHuh]hGm\ h?#\h]h?#\h]h?#6\ h]h@7 h]hGmh?#h]h?#6 h]h?#6bbbc cccdd ee.f/fff)g*gg07$8$H$^`0gd<"$0d^`0a$gd?#0^`0gd<01$^`0gd #1$^#`gdGm ^`gd?# 0^`0gd?#ggChDh9iijj kkkkllOmPmGnHn"$0d1$^`0a$gd01$7$8$H$^`0gd?#"$0d^`0a$gd?#"$0d^`0a$gd< 0^`0gdL 0^`0gd?#OmPmmmGnHnnnoo@pTp^qqqrrrrrrpsssst/ttttt7uKuuuuuuuuuuu v|!h]h?#5:mHnHsH uh]h:Un5;mHnHuh]h@75;mHnHuh]h?#5;mHnHu h]h(h(h]h?#5>*hOchOc6hOc h]hjh?#hh]h?#6 h]h?# h]h0-K+HnoooofpgpqqYrZrrrss.t/ttt\u]uuuuu$a$gd?#"$0d^`0a$gd?# 0^`0gd?# v*v,v.v8v@vAvBvCvLvPv`vavzv~vvvvvvvvvvvv wwww*w+w0wIwJwdwewmw˽tthhZhY0J*CJH*hZhY0J*6CJhY0J*CJhZhY0J*56CJhZhY0J*CJ hZhYhYmHnHuhZhYmHnHuhZhY5mHnHu)h]h?#5;CJ aJ mHnHsH u!h]h?#5:mHnHsH uhY5:mHnHsH u%u,v-v.v8v:vv@vBv$xx$Ifa$gd?# $da$gd?# BvCvLv1& x$Ifgd?#kd$$IfTF47ֈCK"S&)            0Q!44 Faf4ytYTLvPvTvXv\v`v x$Ifgd?#`vavvvzv3* $Ifgd?# $Ifgd?#kd~$$IfTF4ֈCK"S&)      0Q!44 Faf4ytYTzv~vvvv $Ifgd?#vvvv3* $Ifgd?# $Ifgd?#kd_$$IfTF4ֈCK"S&)0Q!44 Faf4ytYTvvvvv $Ifgd?#vvvv3* $Ifgd?# $Ifgd?#kd@$$IfTF4ֈCK"S&)0Q!44 Faf4ytYTvvvvv $Ifgd?#vvv3& n$Ifgd?#kd!$$IfTF4ֈCK"S&)0Q!44 Faf4ytYTvvww ww $Ifgd?#www3& n$Ifgd?#kd$$IfTF4ֈCK"S&)0Q!44 Faf4ytYTwww w%w*w $Ifgd?#*w+w0w3& n$Ifgd?#kd$$IfTF4ֈCK"S&)0Q!44 Faf4ytYT0w5w:w?wDwIw $Ifgd?#IwJwYw3& n$Ifgd?#kd$$IfTF4ֈCK"S&)0Q!44 Faf4ytYTYw]wawfwjwow $Ifgd?#mwnwowpwwwwwwwwwwwyyy!y"y#yHyIyhyȿȷxphx^PDh]h:Un5CJaJh]h?#5:CJaJh]h?#5:h]h:Un5h]h@75h]h?#5h]h?#0J*CJmHnHuh]h?#H*mHnHuh]h?#6mHnHuh]h?#0J*CJhY0J*CJh{'mHnHuh]h?#mHnHuhZhY0J*CJ hZhYhZhYmHnHuhZhYH*mHnHuowpww3$ nx$Ifgd?#kd$$IfTF4ֈCK"S&)0Q!44 Faf4ytYTwwwwww x$Ifgd?#wwwwy3---1$gd?#kdx$$IfTF4ֈCK"S&)      0Q!44 Faf4ytYTyy#yHyIyTyhyiyH}kdg $$Ifl40!     t044 lalyt:Un$xx$Ifa$gd{' $$Ifa$gd?# Pgd?# $ Pa$gd:Un$a$gd?#iyjyyyyyyUJ= V$Ifgd?# x$Ifgd?#kd $$Ifl4F!   ~  $   t0    44 lalyt:Un$xx$Ifa$gd{' $Ifgd?#hyjyyyy zzzzz{e{f{{{{{{{l|||||||||||||||||}}*}+},}5}7}8}N}O}üæüÞ{ jh]hkiUmHnHuh]h?#5:h]h@75h]h?#5 h]hc0Xh]hP!6 h]hP! h]hki h]h?#h]h:UnCJaJh]h:UnCJH*aJhSCJaJh]h:Un5CJaJh]h:UnCJaJ,yyyyyycVI< r$Ifgd:Un &$Ifgd:Un $If^gd?#kd $$IflF! ~ $  t0    44 lalyt:Un V$Ifgd?#yyyyypcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd $$IflF!~ $ t0    44 lalyt:Unyyzz zpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kdy $$IflF!~ $ t0    44 lalyt:Un z zzzzpgZM r$Ifgd:Un &$Ifgd:Un $Ifgd?#kd7 $$IflF!~ $ t0    44 lalyt:Unzzz#z+zpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd $$IflF!~ $ t0    44 lalyt:Un+z,z0z8z?zpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Un?z@zTz\z`zpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kdq$$IflF!~ $ t0    44 lalyt:Un`zazezizmzpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd/$$IflF!~ $ t0    44 lalyt:Unmznz|zzzpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:UnzzzzzpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:UnzzzzzpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kdi$$IflF!~ $ t0    44 lalyt:UnzzzzzpcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd'$$IflF!~ $ t0    44 lalyt:Unzz{{{pgZM r$Ifgd:Un &$Ifgd:Un $Ifgd?#kd$$IflF!~ $ t0    44 lalyt:Un{{{ {'{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Un'{({8{?{F{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kda$$IflF!~ $ t0    44 lalyt:UnF{G{S{Z{a{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Una{b{g{l{s{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Uns{t{{{{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Un{{{{{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kdY$$IflF!~ $ t0    44 lalyt:Un{{{{{pgZM r$Ifgd:Un &$Ifgd:Un $Ifgd?#kd$$IflF!~ $ t0    44 lalyt:Un{{{{{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Un{{{{{pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Un{{ |||pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kdQ$$IflF!~ $ t0    44 lalyt:Un||)|.|3|pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Un3|4|F|J|O|pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:UnO|P|b|g|k|pcVI r$Ifgd:Un &$Ifgd:Un $If^gd?#kd$$IflF!~ $ t0    44 lalyt:Unk|l|||||peWK $$Ifa$gd:Un$x$Ifa$gd?# x$Ifgd?#kdI$$IflF!~ $ t0    44 lalyt:Un|||||||,}-}8}M}N}niiiiiiaaaa$a$gd?#gd?#kd$$IflF! ~ $  t0    44 lalyt:Un O}P}g}h}i}j}k}l}m}n}o}t}}}}}}}}}}}}}}}}}}}}}¹ymymbmbymymWbOh> CJaJh> h> CJaJh(Jh> CJaJh(Jh> 5CJaJh> 5CJaJhYhL5:CJaJh]hL5: h5\h]hL5\ h]h h]hr -h]h?#0J* h]h?#h]h?#5jh]hkiU jh]hkiUmHnHu h]hkijh]hkiUN}l}m}n}o}p}q}r}s}t}}}}}}$xx$Ifa$gd> $1$a$gdL $1$7$8$H$a$gdL H!$dgd~ 1$7$8$H$gd?#dgd?#$a$gd?#}}}}}rd$x$Ifa$gd> x$Ifgd> kdm $$IfTl0PL!     t0644 laytr2T}}}}}th $$Ifa$gd> $Ifgd> kd0!$$IfTl0PL!   t0644 laytr2T}}~~}pd $$Ifa$gd>  $If^ gd> kd!$$IfTl0PL! t0644 laytr2T}~~~~~&~)~*~+~-~A~D~E~F~G~~~~~~~~~~~~HKL.12€ŀƀǀ݀ހ"%&mpq    X[\‚͂hr25CJaJh.CJaJh(Jh> 5CJaJh> 5CJaJh> h> CJaJh> CJaJh(Jh> CJaJH~~&~*~}pd $$Ifa$gd>  $If^ gd> kd"$$IfTl0PL! t0644 laytr2T*~+~A~E~}th $$Ifa$gd> $Ifgd> kdK#$$IfTl0PL! t0644 laytr2TE~F~~~}pd $$Ifa$gd>  $If^ gd> kd#$$IfTl0PL! t0644 laytr2T~~~~}pd $$Ifa$gd>  $If^ gd> kd$$$IfTl0PL! t0644 laytr2T~~~}th $$Ifa$gd> $Ifgd> kdX%$$IfTl0PL! t0644 laytr2THL}pd $$Ifa$gd>  $If^ gd> kd&$$IfTl0PL! t0644 laytr2TLM}pd $$Ifa$gd>  $If^ gd> kd&$$IfTl0PL! t0644 laytr2T}th $$Ifa$gd> $Ifgd> kde'$$IfTl0PL! t0644 laytr2T}th $$Ifa$gd> $Ifgd> kd($$IfTl0PL! t0644 laytr2T.2}th $$Ifa$gd> $Ifgd> kd($$IfTl0PL! t0644 laytr2T23}th $$Ifa$gd> $Ifgd> kdr)$$IfTl0PL! t0644 laytr2T€ƀ}th $$Ifa$gd> $Ifgd> kd!*$$IfTl0PL! t0644 laytr2Tƀǀ݀ހ}th $$Ifa$gd> $Ifgd> kd*$$IfTl0PL! t0644 laytr2Tހ߀"&}th $$Ifa$gd> $Ifgd> kd+$$IfTl0PL! t0644 laytr2T&'mq}th $$Ifa$gd> $Ifgd> kd.,$$IfTl0PL! t0644 laytr2Tqr }th $$Ifa$gd> $Ifgd> kd,$$IfTl0PL! t0644 laytr2T   }th $$Ifa$gd> $Ifgd> kd-$$IfTl0PL! t0644 laytr2T !X\}pd $$Ifa$gd>  $If^ gd> kd;.$$IfTl0PL! t0644 laytr2T\]}pd $$Ifa$gd>  $If^ gd> kd.$$IfTl0PL! t0644 laytr2T‚ÂĂłƂǂȂɂʂ˂̂͂}vvvvvvvvvvvvxgd> kd/$$IfTl0PL!   t0644 laytr2T ͂ۂ6789:=>NPQ]ezƫpY?4hr2>*CJOJQJ2jhhr25>*CJOJQJUmHnHu,jhhr2CJOJQJUmHnHu*jhr2CJOJQJUmHnHtH uhhr2>*CJOJQJ,h%(hr256>*CJOJQJmHnHu5jh%(hr256>*CJOJQJUmHnHu5jhhr256>*CJOJQJUmHnHu hr2hr2hr2h> hr2h> hr2h> 57:;<=OPRSTUVWXYZ[\]^_`abcde $h^ha$gdr2xgd> z  'vn΋Ћڋދ^ЌҌf{&(HJ]| ÷|rh]hD55;h]hL5; hL5; hD55;h]hL6h.h]hLH* h]hLh]hL5h]hL6CJaJh]hL5:\ h5\h]hL5\ hr2hLhr2hL6 hr2hr2hr2hr20J*CJ,e  'vf|]| !gdD5 d`gdLdgdL d1$`gdL d1$gdL1$gdL $1$7$8$H$a$gdLgdLxgd> <gdr2     $ !a$gdD5 !gdD5$a$gd! dh`gd$ !a$gdD5 h5<BVe W]nv6y()W\F 0Lȯ֯ )HϿϿϷϷϫϧϣϘϣϷόόόh?Xhnh`m6h_ h]h`mh`mh5Wh]h!mHnHuh]h!6h]h!5hvVhY h]h!hUhD5 h]hD5h]h!5;h]hD55; hU5;7Šnbֳ xh,UmrdgdU $da$gd d1$gddgd! $da$gd! d1$`gd`m d1$`gd! d`gd! dh`gd!$a$gd!HKհܰJuαױ bֳ*.HTd~h"\^x!=UmhUhU5 h]h h]h h  h]hkihI' h]hhh7hzhhh]h!5;hUh`mhvVh! h]h!h?X>mwm[nqr"GV.7?JPRR!1[fqHIڿ⬨ҤҤҜҤҤҜh2\hShnh^q:hUh!h]hH* h]h h]h!hh h!h!chlh5(hy\ h]h!h?X h 6h?Xh?X6 h?X6;rQR,]^el;dgd[;dgd! $da$gd!gdd d1$`gd+l d1$`gd_)Idgdd 8^8`gd! 8^8`gd d`gd!I"=+,-.Fbfsx7|$8XG'K23\^eǿӻӰӬӻӬӻh]h!5;h]h!H*h+l h]h!ch!c h]hLhThLhdhph5hnhU5h h]hh]hH* h]h! h]hUh7bhzhhhqr6el:Mxz{ *4689:QSuz6?DNu%=>hzhh]h!6h`eah!6 h]hh6h,h`eahj0h!'hZhThSh h]h!h]h!5B;?ME0- ) /dgdL/dgd: /d`gd( /d1$`gd d1$gddgdDH d1$`gd d`gd! d1$`gd,dgd! d`gd!'>?f +-RTlu+wLM"Ckmhl h]hY hDH5 h^5h]hDH5 h]hDHhhh`eah!6h h]h!h]h!5h]h!6 h]h*|hthD_Sh^h`eah`ea6h`eah*|6h`eah*|h!3(1EGLNvw!"5678ддддЛЈr])jV0h]hyCJEHOJQJUaJ+j8M h]hyOJPJQJUVaJ%jh]hyCJOJQJUaJha{CJOJQJaJha{6CJOJQJaJh]h:6CJOJQJaJhCJOJQJaJh]h:CJOJQJaJh]hyCJOJQJaJh ha{hY h]hY 89@ABDXYt}ݺݮݺ݉zdO<%jCRh]h:CJOJQJUaJ)jOh]hyCJEHOJQJUaJ+jgM h]hyOJPJQJUVaJh]hyCJOJQJaJ%jh]hyCJOJQJUaJ"h]h:6CJH*OJQJaJha{CJOJQJaJh]h:6CJOJQJaJ%j2h]h:CJOJQJUaJh]h:CJOJQJaJ%jh]h:CJOJQJUaJ#$78ҭ~k\F+jM h]h(OJPJQJUVaJh]h(CJOJQJaJ%jh]h(CJOJQJUaJha{CJOJQJaJh]h:6CJOJQJaJ%jmh]h:CJOJQJUaJ%jh]h:CJOJQJUaJ"h]hy6CJH*OJQJaJh]hyCJOJQJaJh]hy6CJOJQJaJh]h:CJOJQJaJ89:;BCDFWXuvwxyĵĵpaĵNĵa%jh]h:CJOJQJUaJh]h(CJOJQJaJ"h]h(6CJH*OJQJaJh]h(6CJOJQJaJh]h:6CJOJQJaJ%jh]h:CJOJQJUaJh]h:CJOJQJaJ%jh]h:CJOJQJUaJ%jh]h(CJOJQJUaJ)jh]h(CJEHOJQJUaJ*F|m|[mHm%jh]h:CJOJQJUaJ"h]h(6CJH*OJQJaJh]h(CJOJQJaJh]h(6CJOJQJaJ%jh]h:CJOJQJUaJh]h:CJOJQJaJ%jh]h:CJOJQJUaJ%jh]h(CJOJQJUaJ)jh]h(CJEHOJQJUaJ+j狔M h]h(OJPJQJUVaJFHJLx(*.02@tatOt"h]h(6CJH*OJQJaJ%jh]h:CJOJQJUaJ%jh]h:CJOJQJUaJ)jh]h(CJEHOJQJUaJ+j.M h]h(OJPJQJUVaJh]h(CJOJQJaJ%jh]h(CJOJQJUaJha{CJOJQJaJh]h:CJOJQJaJh]h:6CJOJQJaJ@BDHJr"4$&(,~)*+,iݾݾݯݜݾݾݾݾݐ}gR})j Rh]h(CJEHOJQJUaJ+ǰM h]h(OJPJQJUVaJ%jh]h(CJOJQJUaJh!6CJOJQJaJ%j8h]h:CJOJQJUaJh]h(CJOJQJaJha{CJOJQJaJ%jh]h:CJOJQJUaJh]h:CJOJQJaJ%jWh]h:CJOJQJUaJikry      % ʻʕttٻhUFh]hOCJOJQJaJ%jh]hOCJOJQJUaJhLCJOJQJaJhOeCJOJQJaJ)jgUh]hYCJEHOJQJUaJ%j"M hYOJPJQJUVaJ%jh]h(CJOJQJUaJh]h]CJOJQJaJh]h(CJOJQJaJh!6CJOJQJaJh]h:CJOJQJaJhCJOJQJaJ% & ' ( ) / 0 1 K O z }         ` l           ĵ{ococTh]h:CJOJQJaJhOeCJOJQJaJhkCJOJQJaJhCJOJQJaJ"hbhhbh6CJH*OJQJaJhbhhbh6CJOJQJaJhbhCJOJQJaJh]hOCJOJQJaJ%jh]hOCJOJQJUaJ)j/YhLhbhCJEHOJQJUaJ%ji#M hbhOJPJQJUVaJ)    20 Jt 8% d1$gdOe $da$gdOe $da$gd~dgd!dgdY/dgd: /d`gdL/dgd] /d`gdO/dgdbh             & ( * . @      $ ( , 6 :    οίο{h{\{P{P{P{{\{P{hOeCJOJQJaJhkCJOJQJaJ%j;]h]h:CJOJQJUaJh]h:CJOJQJaJ%jh]h:CJOJQJUaJ"h]h]6CJH*OJQJaJh]h]6CJOJQJaJh]hOCJOJQJaJh]h]CJOJQJaJ"h]h]6CJH*OJQJaJh]hO6CJOJQJaJ           *,.02忪埐}l]QA]h]h]CJH*OJQJaJh]CJOJQJaJh]h:CJOJQJaJ!jh]hbhEHOJQJU%jF%M hbhOJPJQJUVaJjh]h]OJQJUh]h]OJQJ)jx|h]hbhCJEHOJQJUaJ%j7%M hbhOJPJQJUVaJ%jh]h]CJOJQJUaJh]h]CJOJQJaJh!6CJOJQJaJ.4dhz맛nYnnF%jh]h:CJOJQJUaJ(jh]h]6CJOJQJUaJ(jh]h]6CJOJQJUaJhOeCJOJQJaJhkCJOJQJaJhbhCJOJQJaJh]h:CJOJQJaJh]h]6CJOJQJaJ(jh]h:6CJOJQJUaJh]h:6CJOJQJaJ(jh]h:6CJOJQJUaJprݾwk\Gwk(jph]hOe6CJOJQJUaJhOehOeCJOJQJaJhCJOJQJaJhOe6CJOJQJaJ(j{h]hOe6CJOJQJUaJh]hOe6CJOJQJaJ(jh]hOe6CJOJQJUaJhOeCJOJQJaJ%jh]h:CJOJQJUaJh]h:CJOJQJaJ%jh]h:CJOJQJUaJ "$&(*.02P빭빔p`pTHh:CJOJQJaJh]CJOJQJaJh]h:6CJOJQJaJh]h:CJOJQJaJ(jZh]hOe6CJOJQJUaJhOeCJOJQJaJhOe6CJOJQJaJhCJOJQJaJh6CJOJQJaJ(jeh]hOe6CJOJQJUaJh]hOe6CJOJQJaJ(jh]hOe6CJOJQJUaJ,-./67ܹ̕yfQ̕AhVhV6CJOJQJaJ)j3hVhVCJEHOJQJUaJ%jM hVOJPJQJUVaJh:CJOJQJaJh]h:6CJOJQJaJh]h:CJOJQJaJ)jO0hVhVCJEHOJQJUaJ%jWM hVOJPJQJUVaJjhVCJOJQJUaJhVCJOJQJaJh]CJOJQJaJh!6CJOJQJaJ789:ABCEij˼˼˼˼znbVJh CJOJQJaJhVCJOJQJaJh!6CJOJQJaJh:CJOJQJaJhCJOJQJaJ%jQh]h:CJOJQJUaJh]h:6CJOJQJaJ%j7h]h:CJOJQJUaJh]h:CJOJQJaJ%jh]h:CJOJQJUaJhVhV6CJOJQJaJ"hVhV6CJH*OJQJaJ    '<BMo,24R[g?z=[fлﬤ}yuyqj} h]h~h!h}6hhh h~h,[~hOehd9h]h!6 h]h!hYhY5h]h:CJOJQJaJ)jlhVh W@CJEHOJQJUaJ%jM h W@OJPJQJUVaJhVCJOJQJaJjhVCJOJQJUaJ)fq89v .NJLbdfhrt@.6brvxY r s t    "" "ܼ츴참ܜܸh8hH*hKOhkhk6h8h7ash^ehzMyhh6h8hfH*hfh8h6'H*hhOeh \hhkh6'h}6h~h= """""0#j###$($$"%&%(%2%6%8%H%^%`%%f&&&'M'W'['~'''((((()))*[+--. .'.G...>/I//00쮲앍썕h,hbhh[;h24hh5h]h5 h5h h]h7hjh \hh8 h]h!h]h!5 h5 h~hOeh^eh6hzMyhh6h^eh68%^%().04748;>>@CF d1$`gdr2 d1$`gdbh d1$`gdt d`gdQp d1$gddgd) d`gdzMy d1$`gd[; d`gd24dgd! $da$gd!000001 21222;2333344464748445566t6z666677888:;;K<N<w<<<<<<<==<>>>>>>>>>>???0???@@Ažɾžžݯūhqpshbhh]h!5 hvzTh) h]hvzTh GhthQph)h)h)5 h)5h! h]hzMyhvzThzMyhj h]h!BAAbB{BBBCCCCCC5D>D~DDFFFFFFFFFGGG G̾}p}pcSh'@hh{5mHnHsH uhZN5mHnHsH uht5mHnHsH uh'@h5mHnHsH u)h]hZN5;CJ aJ mHnHsH u!h]hZN5:mHnHsH uht5:mHnHsH uh`~&5:mHnHsH uh]h:5 hZN5hO h]h!h Ghh[;hhp)hqpshtFFFFG GGGGG!G$G(G+G/G2G6G?GDGGG $$Ifa$gdz$xx$Ifa$gdZN hd^hgd{ & F hhd^hgd{ $da$gdZN$a$gdZN GGGG G!G'G(G.G/G5G>G?GCGFGGGHGVG^G_GgGhGpGqGyGzGGGGGGGGGGGGGGGGGGGGGGGHHHH H HHHHHHHHƼƼƼƼƼƼƱƼƼƼƼƼƼƱƧhPHh'@h0J*CJaJh'@h0J*CJaJhPHhzCJaJhz0J*CJaJhPHhz0J*CJaJhPHhz5CJaJhz5CJaJmHnHu"hPHhz5CJaJmHnHu!z              0!44 Faf4yt TVG[G_GdGhGmGqGvGzGGGGG $Ifgdbh x$Ifgdbh $Ifgdbh x$Ifgdbh GGG  $Ifgdakd|o$$IfTF4֞4 Ld|>!z       0!44 Faf4yt TGGGGGGGGGGGGG $Ifgdz $Ifgdz GGG  $Ifgdakdvp$$IfTF4֞4 Ld|>!z 0!44 Faf4yt TGGHH HHHH H%H*H/H4H $Ifgdz $Ifgdz H H$H%H)H*H.H3H4H5HSHWHXH\H]HaHbHfHgHkHlHpHqHuHvHzH{HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHIIIIII2I3I=IBICIDIeIfIpIuIvIwIhPHhzCJaJhPHhz0J*CJaJhz0J*CJaJh'@h0J*CJaJhPHh'@h0J*CJaJM4H5HSH  $Ifgdakdbq$$IfTF4֞4 Ld|>!z 0!44 Faf4yt TSHXH]HbHgHlHqHvH{HHHHH $Ifgdz $Ifgdz HHH  $IfgdakdNr$$IfTF4֞4 Ld|>!z 0!44 Faf4yt THHHHHHHHHHHHH $Ifgdz $Ifgdz HHI  $Ifgdakd:s$$IfTF4֞4 Ld|>!z 0!44 Faf4yt TII I I I IIII4I5I6I7I8I9I>ICIDIgIhIiI x$Ifgd  nx$Ifgd Ffxx n$IfgdaFfQu $Ifgdz $IfgdaiIjIkIlIqIvIwIxIIIJJJJJJJJ$xx$Ifa$gd'@hgdZN & F hhd^hgd{1$gdz1$gdZNFf{ x$Ifgd $Ifgd  x$Ifgd wI|I}IIIIIIIII$J)JJJJJJJJJJJʿ⤜|lgUI<h]h-0J*CJaJh{h-5CJaJ"h{h-5CJaJmHnHu h'@h5h{hZN5mHnHsH uh{h{5mHnHsH uh{h'@h5mHnHsH uh-0J*CJhZN0J*CJhzhPH0J*CJhPH0J*CJh]hZN0J*CJh`,c0J*CJh]hz0J*CJmHnHuhz0J*CJh{mHnHuh]hZNmHnHuJJJJJJJJJJJJJJJJJJJJJ n$Ifgd- $Ifgd'@hFfҁ x$Ifgd'@h x$Ifgd'@hFf~$xx$Ifa$gd'@hJJJJJJJJJJJJKKKKKKK#K$K)K,K:K;K>K?KBKCKFKMKNKRKVKdKeKhKiKlKmKpKqKtKzK{KKKKKKKKKKKKKKKKKKKKKKKKKKKKKLLܼh-h-0J*CJh{0J*CJaJh]h-CJaJh-h-0J*CJaJh]h-0J*CJaJh-0J*CJaJJJJJJJJJJJJKKKKKK K!K"K#K$K;K?KFf n$Ifgd- $Ifgd'@hFf n$Ifgd- $Ifgd-?KCKGKHKIKJKKKLKMKNKeKiKmKqKuKvKwKxKyKzK{KKK n$Ifgd-Ffb n$Ifgd'@hFf> n$Ifgd-KKKKKKKKKKKKKKKKKKKKKLLFf n$Ifgd- $Ifgd-Ff n$Ifgd{ Z$Ifgd-LLL L!L$L%L(L)L,L-L0L1L4L7L8LZL[L^L_LbLcLfLgLjLkLnLoLrLsLvLwLzL|L}LLLLLLLLLLLLLLLLLLLLLLLLLLLLLLLԶh[h-h-6hVh- h<'_5CJOJQJ\^JaJh]h-0J*CJaJh]h-CJaJh{0J*CJaJh-h-0J*CJaJh-0J*CJaJ@L!L%L)L-L1L5L6L7L8L[L_LcLgLkLoLsLwL{L|L}LLL nx$IfgdV nx$IfgdVFf $Ifgd-FfΗ n$Ifgd-LLLLLLLLLLLLLM*M+M?MGMOMWM $$Ifa$gd3gdZN$a$gdogd- 7$8$H$gd<'_Ff$ nx$IfgdVLLMM*M+M>MXMbMcMhMsMtMuMvMyM|M}MMMMMMMMMMMMMMMMMMĹ{ochwh36CJaJhwhGy5CJaJhwh&CJaJh*|CJaJhGyCJaJh&CJaJh 7tCJaJhCJaJhwhwCJaJhwhGyCJaJhwhGy6CJaJhwh35CJaJhwhw5CJaJh`~&h35: h35 ho5h"WMXMbMlMsM}MMMMuhhhhhh V$IfgdGy  $If^ gd&|kdK$$Ifl\T!  6  6  6   t644 lalytD_SMMMMMMMMMuhh[[[[ V$Ifgdw V$IfgdGy  $If^ gd&|kd$$Ifl\T! 6 6 6  t644 lalytD_SMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMN N NNNNNNNNNNNڿڷڿڷnjh7CJaJhwh36CJaJhwhGy6CJaJhwhDCJaJhDCJaJh3CJaJh*|CJaJhwhwCJaJh 7tCJaJhCJaJhwhGyCJaJhwh3CJaJ7MMMMMMMMMuhhhhhh V$IfgdGy  $If^ gd&|kd$$Ifl\T!666 t644 lalytD_SMMMMNNNN%Nuhhhhhh V$IfgdGy  $If^ gd&|kdn$$Ifl\T!666 t644 lalytD_SN N#N$N%N&N(N)N2N3N:N>N@NCNDNINKNMNPNQNUNWNXNYNZNeNuNvNwN{N|N}NNNNNNNNNNNNNƺүҧҏyҧүnyҧhD_ShGyCJaJhD_Sh3CJaJhD_ShD_SCJaJhD_SCJaJh3CJaJh 7tCJaJhCJaJhwhwCJaJhwh36CJaJhwhGy6CJaJhwh3CJaJhwhDCJaJhDCJaJh7CJaJh*|CJaJ+%N&N)N3N:NDNKNQNYNuhhhhhh V$IfgdGy  $If^ gd&|kd$$Ifl\T!666 t644 lalytD_SYNZNvNwN}NNNNuhhhhh V$IfgdGy  $If^ gd&|kdȣ$$Ifl\T!666 t644 lalytD_SNNNNNNNNuhhhhh V$IfgdGy  $If^ gd&|kdu$$Ifl\T!666 t644 lalytD_SNNNNNNNNNNNNNNNNNNNOO O O O O*O+O,O.O0O3O4O7O9O;O=O?OBOCOFOHOIOJOKOLOhOiOjOnOoOpOtOuOwOxOzO|O}OʷҜʜœʜ푯hD_ShGyCJaJhDCJaJhD_Sh3CJaJhD_SCJaJhwhDCJaJh3CJaJh7CJaJh*|CJaJhwhwCJaJhCJaJhwh3CJaJh 7tCJaJ9NNNNNNO Ouhhhhh V$IfgdGy  $If^ gd&|kd"$$Ifl\T!666 t644 lalytD_S O O+O,O4O;OCOJOuhhhhh V$IfgdGy  $If^ gd&|kdϥ$$Ifl\T!666 t644 lalytD_SJOKOiOjOpOwO}OOuhhhhh V$IfgdGy  $If^ gd&|kd|$$Ifl\T!666 t644 lalytD_S}OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO P PPP&PǿҬҙǿҬҬǿҙҙҿh3CJaJhwhwCJaJh 7tCJaJhCJaJhD_Sh3CJaJhD_SCJaJhD_ShGyCJaJhwh3CJaJhwhDCJaJhDCJaJh7CJaJh*|CJaJ9OOOOOOOOuhhhhh V$IfgdGy  $If^ gd&|kd)$$Ifl\T!666 t644 lalytD_SOOOOOOOOuhhhhh V$IfgdGy  $If^ gd&|kd֧$$Ifl\T!666 t644 lalytD_SOOOOOOO Puhhhhh V$IfgdGy  $If^ gd&|kd$$Ifl\T!666 t644 lalytD_S PP(P*P,P>PLPuhhhh V$IfgdGy  $If^ gdD_S|kd0$$Ifl\T!666 t644 lalytD_S&P(P,P0P6PPFPHPJPLPNPtPvP~PPPPPPPPPPPPPPPPPPPPPPPQDQFQJQNQPQRQXQZQ`QdQҿ귬ҍҿڅz귬hwh}U|CJaJh}U|CJaJhGyCJaJhwhGyCJaJhD_ShGyCJH*aJhD_ShGyCJaJhD_SCJaJhwhDCJaJh3CJaJhDCJaJh7CJaJh*|CJaJhwh3CJaJhD_Sh3CJaJ0LPNPPPPPPuhhhh V$IfgdGy  $If^ gd&|kdݩ$$Ifl\T!666 t644 lalytD_SPPPPPPPuhhhh V$IfgdGy  $If^ gd&|kd$$Ifl\T!666 t644 lalytD_SPPFQHQJQZQhQuhhhh V$IfgdGy  $If^ gd&|kd7$$Ifl\T!666 t644 lalytD_SdQfQhQjQQQQQQQQQQQQQQQRRR2R4R$$Ifl\T!666 t644 lalytD_SRRRRRRuiii $$Ifa$gdw  $If^ gd}U||kd$$Ifl\T!666 t644 lalytD_SRRRRRRRRRRRRRSSS SSS S"S&S.S0S4S8S:SSuiii $$Ifa$gdw  $If^ gd}U||kdE$$Ifl\T!666 t644 lalytD_S>S@SHSVSdSrSuiii $$Ifa$gdw  $If^ gd}U||kd$$Ifl\T!666 t644 lalytD_SrStSSTTT-T.TKTLTMT}}}uuum}mdgdZN$a$gdZNgdZN|kd$$Ifl\T! 6 6 6  t644 lalytD_S .T/TFTGTHTITJTKTLTMTVTWTXTTTTUUUUUVVVVVVɿ펂zoe\UNDh]h70J$6 hZN0J$6 h70J$6h]hZN0J$h]hZN0J$6h]h7mH sH hZNmH sH h]hZN6mH sH h]hZNmH sH h]hZN6h!1hZN5;\h!15;\h!1ho5;\h]hZN5>*h]hZN5jZh]h`~&U jh]hZNUmHnHu h]hZNjh]hZNUMTXTTUUUVVWW8X9XXXVYWY'Z(ZZ 0^`0gd7 0^`0gd U-"$0d^`0a$gd U-#0d1$^`0gd U-01$7$8$H$^`0gd U-$d7$8$H$a$gd!1VVcWWWWWWW X'X7X8X9XZZX[Y[[[\&]'] ^ ^^^V_W_````_a 0^`0gd 1$7$8$H$gd7 0[$\$^`0gd U- 0^`0gd^q:%0d^`0gd U- 0^`0gd U-_``````````0aLa_a`a3b^bobpbqbb c2c3c4chcccccc d d dddddd>e^eoepeqeeeffff컯htmHnHuh]h7mHnHuhZNmHnHuh]hZN6mHnHuh]hZNmHnHu h]hhEhE6hEhh]h6 h]hh7 h]h7hZN h]hZNh]hZN61_a`apbqb3c4ccc d dddpeqeffffgghh 0^`0gd U-01$^`0gd07$8$H$^`0gd U-"$0d^`0a$gd U-fffggggg_hshhhhiiiJiSixiij1jDjkk&k'kTkakƺod`X`QIQ`h]hz6 h]hzhIhz6hzhzmHnHsH uhNchz6mHnHsH uhNchzmHnHsH uhNchzmH sH hzmH sH  h]hhZNh]hZN6 h]hZNh]h7mHnHuhZNmHnHuh]hZN6mHnHuh]hZNmHnHuhtmHnHuhtht6mHnHuhiiDjEj&k'kkkllllcmdmmlnmno 1$7$8$H$gd"$0d^`0a$gd U-01$^`0gdzgd 0^`0gdz 0^`0gd U-#0d1$^`0gdzakkklllll;mRmdmmmmHn\nnnogo{ooop'p(p)pppqqqqqFrcrwrrrrrrr)sEswp h]h:htht6hth]hZN6] h]h U-h]hz6 h]hz hhhhhhB*phhhhh6B*]phhhhhB*phhh6hh]h5(6 h]h5( h]hhZNh]hZN6 h]hZN,oooo(p)pqqqqvrwrrrXsYsss#0d1$^`0gd U-"$0d^`0a$gd U-"$0d^`0a$gdz h^h`gd U- 0^`0gd U-01$7$8$H$^`0gd U-EsWsXsYs\ssssssttu u u%uhuuuuuvvvvvvvvv(w)w*w9wwwww[xwxxxyyպunn h]hh]h U-mHnHuhZNmHnHuh]hZN6mHnHuh]hZNmHnHuh]hZN6h]h U-mHnHsH uhZNmHnHsH uh]hZN6mHnHsH uh]hZNmHnHsH uh:h:mH sH  h]h U-hZN h]hZN+s u uuuvvvv)w*wwwxxyyyyzzr{s{gd07$8$H$^`0gd U-01$^`0gdl01$^`0gd 0^`0gd U-yyyyyyyy\zxzzzM{a{q{r{s{{|j|k|l|y|||<}=}>}}}}{~~+?GLMZ}~浥zha{h!16CJOJQJaJha{h!1CJOJQJaJh!1CJOJQJaJhh!15CJOJQJaJ h!15\hhh6h h]hhlhl6hl h]hlhZNh]hZN6 h]hZN h]h U-*s{k|l|=}>}}}~~MZ}~ d/dgd!1 /d1$`gd!1 /$1$a$gd!1 $7$8$H$a$gd!101$^`0gd01$^`0gd 0^`0gd U-"$0d^`0a$gd U-$&ր<܁ށ÷èjZH"h]h!16CJH*OJQJaJh]h!16CJOJQJaJ)jh]h!1CJEHOJQJUaJ+jM h]h!1OJPJQJUVaJ%jh]h!1CJOJQJUaJh]h!1CJOJQJaJh!6CJOJQJaJh!1CJOJQJaJha{h!16CJOJQJaJha{h!1CJOJQJaJ"ha{h!16CJH*OJQJaJ *+-./678:]^noʂݺݕݺݺvʺ݉݉cʺ%jh]h!1CJOJQJUaJ%j~h]h!1CJOJQJUaJh!1CJOJQJaJ%jh]h!1CJOJQJUaJ"h]h!16CJH*OJQJaJh]h!16CJOJQJaJ%jh]h!1CJOJQJUaJh]h!1CJOJQJaJ%jh]h!1CJOJQJUaJ&  "&046\^`bdftvx|ƄЄ҄||iT)jK[hahaCJEHOJQJUaJ%jM haOJPJQJUVaJh!1CJOJQJaJ%j""h]h!1CJOJQJUaJ)jh]h!1CJEHOJQJUaJ+jM h]h!1OJPJQJUVaJh!6CJOJQJaJhEHCJOJQJaJ%jh]h!1CJOJQJUaJh]h!1CJOJQJaJ H̅9`Ԇ0123456789:;<=>?@A$a$gd!1/dgd!1.@DFJNXZąƅȅʅօ؅څ姒jZH"h]h!16CJH*OJQJaJh]h!16CJOJQJaJ)jbh]haCJEHOJQJUaJ%jM haOJPJQJUVaJ)j^hahaCJEHOJQJUaJ%jM haOJPJQJUVaJ%jh]h!1CJOJQJUaJh!6CJOJQJaJh!1CJOJQJaJh]h!1CJOJQJaJhEHCJOJQJaJ 9:<>HI\]^_ݺݕ݆wl]lJ9]!jh]haEHOJQJU%jM haOJPJQJUVaJjh]h!1OJQJUh]h!1OJQJh!6h!6CJOJQJaJh!6h!1CJOJQJaJ%jSh]h!1CJOJQJUaJ"h]h!16CJH*OJQJaJh]h!16CJOJQJaJ%jh]h!1CJOJQJUaJh]h!1CJOJQJaJ%j%fh]h!1CJOJQJUaJ_`kІц҆ӆ,nY)jNh]haCJEHOJQJUaJ%jM haOJPJQJUVaJh!6CJOJQJaJha{h!16CJOJQJaJ%jh]h!1CJOJQJUaJ%jh]h!1CJOJQJUaJ"h]h!16CJH*OJQJaJh]h!16CJOJQJaJh!1CJOJQJaJh]h!1CJOJQJaJ,-./0abuv02؈ڈ.024ĵ}q}V;5jh%(hr256>*CJOJQJUmHnHu5jhhr256>*CJOJQJUmHnHuhr2hr26CJaJhr2CJaJhr2hr2CJaJhr2hr25CJaJhr2hr25CJaJh]h!15 h!15h]h!1CJOJQJaJ%jh]h!1CJOJQJUaJ)jɿhahaCJEHOJQJUaJ%jM haOJPJQJUVaJABCDEFGHIJKLMNOPQRSTUbuv2ڈ $Ifgdr2$a$gdr2$a$gd!1ڈ068:<`bfhjlnprtvxz|~$h$If^ha$gdr2$If^`gdr2 $Ifgdr246<>^bd|'(ˋ̋wxyzŮ۔ۉ}rerYPrYrYPrYrH}rhr2CJaJhr26CJaJhr2hr26CJaJhr2hr20J*CJaJhr2hr2CJaJhr2hr25CJaJhr2>*CJOJQJ2jhhr25>*CJOJQJUmHnHu,jhhr2CJOJQJUmHnHu*jhr2CJOJQJUmHnHtH uhhr2>*CJOJQJ,h%(hr256>*CJOJQJmHnHu(qAxyz{jkd$$Ifl!! t0!644 lalyt!1 $Ifgdr2$h$If^ha$gdr2 zJ :BDFHJ~8:r\b>@BD<bƢX`t~¤Ĥҥب 輰hr2h!16CJaJhr2hECJaJhr2hr2B*CJaJphh!1B*CJaJphhr2h!1B*CJaJphh!1CJaJhr2hr2CJaJUhr2CJaJhr2h!1CJaJhr2h!15CJaJ6z ƑdޒN D~jkdf$$Ifl!! t0!644 lalyt!1  $Ifgd!1 $Ifgd!1 = .92/.93/.94/.93/.93 for the iPhone,  = .90/.92/.93/.93/.93 for the current brand) 1. What I get from [brand] is worth the costs. 2. All things considered (price, time, and effort), (brand) is a good buy in the (category). 3. Compared with other (category) brands, [brand] is good value for the money. 4. When I use [brand], I feel I am getting my money s worth.Satisfaction with the cellular phone service provider (7-point Likert, strongly disagree/strongly agree.  = .89/.89/.89. The last two waves used only Item 2) 1. The service provider for my cell phone offers outstanding service quality. 2. I am very satisfied with the service provider for my cell-phone. 3. The service provider for my cell phone handles all of my complaints in a satisfactory manner. Economic switching costs (Jones et al. 2007; Burnham et al. 2003; 7-point Likert, strongly disagree/strongly agree.  = .85/.86/.85/.78/.71) 1. Staying with this brand of cell phone allows me to get discounts and special deals. 2. Staying with this brand of cell phone saves me money. 3. Staying with this brand of cell phone allows me to get extra benefits. 4. Switching to a new brand of cell phone will probably involve hidden costs/charges. 5. I am likely to end up with a bad deal financially if I switch to a new brand of cell phone. Procedural switching costs (Jones et al. 2007; Burnham et al. 2003; 7-point Likert, strongly disagree/strongly agree.  = .81/.81/.85/.78/.77) 1. If I switched to another brand of cell phone, I might have to learn new routines and ways of using a new cell phone. 2. If I switched to another brand of cell phone, it might be a real hassle. 3. If I switched to another brand of cell phone, I might have to spend a lot of time finding a new cell phone. 4. I cannot afford the time to get the information to fully evaluate other brands of cell phone. 5. There are a lot of formalities involved in switching to a new brand of cell phone. Consumer innate innovativeness (adapted from Steenkamp and Gielens 2003; 7-point Likert, strongly disagree/strongly agree.  = .79) 1. In general, I am among the first to buy new products when they appear on the market. 2. I enjoy taking chances in buying new products. 3. I am usually among the first to try new brands. 4. When I see a new product on the shelf, I m reluctant to give it a try. (r) 5. I am very cautious in trying new products. (r) 6. I rarely buy brands about which I am uncertain how they will perform. (r) 7. If I like a brand, I rarely switch from it just to try something new. (r) 8. I do not like to buy a new product before other people do. (r) (r) Reverse coded items. Susceptibility to Social Influence (Bearden, Netemeyer, and Teel 1989; 7-point Likert, strongly disagree/strongly agree). Information ( = .88) 1. I often consult other people to help choose the best alternative available from a product class. 2. To make sure I buy the right product or brand, I often observe what others are buying and using. 3. If I have little experience with a product, I often ask my friends about the product. 4. I frequently gather information from friends or family about a product before I buy. Purchasing norms ( = .95) 1. If I want to be like someone, I often try to buy the same brands that they buy. 2. It is important that others like the products and brands I buy. 3. I rarely purchase the latest fashion styles until I am sure my friends approve of them. 4. I often identify with other people by purchasing the same products and brands they purchase. 5. When buying products, I generally purchase those brands that I think others will approve of. 6. I like to know what brands and products make good impressions on others. 7. If other people can see me using a product, I often purchase the brand they expect me to buy. 8. I achieve a sense of belonging by purchasing the same products and brands that others purchase.  It is possible that CBI and perceived value interact with one another. This is in line with theorization that marketing exchanges might involve both utilitarian and symbolic aspects, so called mixed exchange (Bagozzi 1975). For example, one may argue that brands that deliver high perceived value might make the customer-brand relationship more rewarding, hence more salient. I tested this interaction by adding another term into the Level-1 regression models, but found no evidence to this effect. This suggested that CBI and value jointly predict customer behaviors in an additive rather than interactive manner. In addition, none of the interaction terms between national culture and the square terms of the focal predictors was significant.     PAGE   PAGE iii  PAGE x PAGE  PAGE   PAGE 39 PAGE   PAGE 48 PAGE  PAGE 49 PAGE  PAGE 61 PAGE  PAGE 90 Level 1 Level 2 My [brand] s Identity Identity A Customer-Brand Relationship Drivers Relative Strength CBI Saliency (-) Relative perceived value (-) My [brand] s Identity Identity B CATEGORY w/ EFFORT Logit  ADDIN EN.CITE <EndNote><Cite><Author>Gonul</Author><Year>2001</Year><RecNum>9</RecNum><record><rec-number>9</rec-number><ref-type name="Journal Article">17</ref-type><contributors><authors><author>Gonul, Fusun F.</author><author>Carter, Franklin</author><author>Petrova, Elina</author><author>Srinivasan, Kannan</author></authors></contributors><titles><title>Promotion of Prescription Drugs and Its Impact on Physicians&apos; Choice Behavior</title><secondary-title>Journal of Marketing</secondary-title></titles><periodical><full-title>Journal of Marketing</full-title></periodical><pages>79-90</pages><volume>65</volume><number>3</number><keywords><keyword>PHYSICIANS</keyword><keyword>MEDICINE -- Formulae, receipts, prescriptions</keyword><keyword>PATIENTS</keyword><keyword>DRUGS -- Marketing</keyword></keywords><dates><year>2001</year></dates><publisher>American Marketing Association</publisher><urls><related-urls><url>http://ezproxy.lib.uh.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&amp;db=bth&amp;AN=4856485&amp;site=ehost-live </url></related-urls></urls></record></Cite></EndNote>(Gonul et al. 2001) Hidden Markov  ADDIN EN.CITE <EndNote><Cite><Author>Moon</Author><Year>2007</Year><RecNum>83</RecNum><record><rec-number>83</rec-number><ref-type name="Journal Article">17</ref-type><contributors><authors><author>Moon, Sangkil</author><author>Kamakura, Wagner A.</author><author>Ledolter, Johannes</author></authors></contributors><titles><title>Estimating Promotion Response When Competitive Promotions Are Unobservable</title><secondary-title>Journal of Marketing Research</secondary-title></titles><periodical><full-title>Journal of Marketing Research</full-title></periodical><pages>503-515</pages><volume>44</volume><number>3</number><keywords><keyword>MARKETING research</keyword><keyword>SALES promotion</keyword><keyword>MARKETING -- Mathematical models</keyword><keyword>SPECIAL events -- Marketing</keyword><keyword>CONSUMERS</keyword><keyword>PSYCHOLOGY</keyword><keyword>MARKOV processes</keyword><keyword>MONTE Carlo method</keyword><keyword>MATHEMATICAL models</keyword><keyword>hidden Markov model</keyword><keyword>hierarchical Bayes analysis</keyword><keyword>missing data problem</keyword><keyword>sales forecasting</keyword><keyword>sales promotion</keyword></keywords><dates><year>2007</year></dates><publisher>American Marketing Association</publisher><urls><related-urls><url>http://ezproxy.lib.uh.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&amp;db=bth&amp;AN=25685153&amp;site=ehost-live </url></related-urls></urls></record></Cite></EndNote>(Moon, Kamakura, and Ledolter 2007) NBD  ADDIN EN.CITE <EndNote><Cite><Author>Venkataraman</Author><Year>2007</Year><RecNum>81</RecNum><record><rec-number>81</rec-number><ref-type name="Journal Article">17</ref-type><contributors><authors><author>Venkataraman, Sriram</author><author>Stremersch, Stefan</author></authors></contributors><titles><title>The Debate on Influencing Doctors&apos; Decisions: Are Drug Characteristics the Missing Link?</title><secondary-title>Management Science</secondary-title></titles><periodical><full-title>Management Science</full-title></periodical><pages>1688-1701</pages><volume>53</volume><number>11</number><keywords><keyword>PHYSICIANS</keyword><keyword>DECISION making</keyword><keyword>MARKETING</keyword><keyword>DRUGS -- Side effects</keyword><keyword>POLITICAL planning</keyword><keyword>SAMPLING</keyword></keywords><dates><year>2007</year></dates><publisher>INFORMS: Institute for Operations Research</publisher><urls><related-urls><url>http://ezproxy.lib.uh.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&amp;db=bth&amp;AN=27875318&amp;site=ehost-live </url></related-urls></urls></record></Cite></EndNote>(Venkataraman and Stremersch 2007) Behavioral Intentions Behavior under normal conditions Repurchase intention Behavior under abnormal conditions Customer forgiveness Perceived Value Close Together but Separate Very Small Overlap Small Overlap Moderate Overlap Large Overlap Very Large Overlap Complete Overlap Far Apart H G F E D C B Close Together but Separate Very Small Overlap Small Overlap Moderate Overlap Large Overlap Very Large Overlap Complete Overlap Far Apart H National culture Individualism/Collectivism Uncertainty avoidance Control variables Brand Trust Socio-demographic variables Product categories Control variables Consumer characteristics Innate innovativeness (+) Susceptibility to social influence (information and purchasing norms) Gender, Age Consumer satisfaction with service provider Switching costs to the new brand Economic switching cost (-) Procedural switching cost (-) Switching to the new brand A D Customer-Brand Identification E C G F  >(֗HܘFH $Ifgd!1jkd$$Ifl!! t0!644 lalyt!1 HJ:l\ҝ@B $Ifgd!1jkd*$$Ifl!! t0!644 lalyt!1 BDNbȡdȢb $Ifgd!1jkd$$Ifl!! t0!644 lalyt!1¤Ĥ zz`S $If]gd!1 & F VV$If^V`gdE0$If^`0gd!1jkd$$Ifl!! t0!644 lalyt!1 $Ifgd!1 v(ب:pʭ $If]gd!1 & F VV$If^VgdE $Ifgd!1 jlnrtxz~1$gdQy1$gdr2jkdP$$Ifl!! t0!644 lalyt!1 dhjlnptvz|´δдҴԴشڴ "$0248:FHLNRT`bdhýýýòýçýýýýýçýýýçýýh:0JmHnHuh00JmHnHu h00Jjh00JUjh0Uh%+ h]h0h0jh00JU h]h!1hr2h!1CJaJBִش  $h]ha$gdNh]hgd?# &`gd?#h]hgd?# &`#$gd?#$a$gd+ &`#$gd+ "468PRdfhȵ$&`#$a$gdLh]hgd! &`#$gdLh]hgdS &`#$gd_ $h]ha$gdP!h]hgd?# &`gd?#hjvx|~ĵƵȵ̵εڵܵ$&HL ྴhhh0CJaJ hr2h0h05B*hphhr2h05B*hphh0B*hphhr2h0B*hphhSih056 h056h%+h0h:0JmHnHu h00Jjh00JU2ȵʵ̵$&gdr2`gdr2gdki &`#$gdLh]hgd!  ~J`gdr2 7$8$H$gdZN & F7$8$H$gdZN $7$8$H$a$gdZN $7$8$H$a$gd3gdr2 ~ DJ¸ĸȸʸɻ|q||q|qhnCuh0CJaJh0CJaJjh05U h05hhh0CJaJhh0CJaJ hh0h05B*hphhJh05B*hphh0B*hphh0B*phhvA=h0B*phh05B*\phh0hh0CJaJ,¸ĸƸȸʸ&(j $7$8$H$a$gdki h^h`gdnCugdj h^h`gd^gdr2$&(8hjl02jlnܱtܱtj^OC^h0B*CJ hphh_h0B*CJhphh0B*CJhphh\h05\ hSih0B*CJ\aJphh0B*CJ\aJphhSih0B*CJaJphh05B*CJ\aJph#hSih05B*CJ\aJphhB1Bh05B*\phh05B*\phh0h*th0CJaJhnCuh0CJaJjh05Uj024lnpgdr2 $7$8$H$a$gdki 7$8$H$gdki & Fxx7$8$H$gdki&(*PRVX\^bdvxz|~ʾٲ٨٨|t||h0CJaJhh0CJaJ hh0hh0CJaJhhh0CJaJh0h0B*hphh0B*CJ hphh0B*CJhphhmh0B*CJhphh0B*CJhphh0B*CJ hphhJh0B*CJhph-(*,.PRTVXZ\^`bdxz|~gdr2gdr2$&(NPRnpr澯yyyhmh0B*CJhphh0B*CJ hphhJh0B*CJhphh0B*CJ hphh_h0B*CJhphh0B*CJhphhhh0CJaJ hhh0h0B*hphh0hh0CJaJhh0CJaJ*&(*PRTprtgdr2  02468:<>@`bjlnÿâvncvvnch"h0B*phh0B*phh"h05B*\phh05B*\phh0CJaJhh0CJaJ hh0hh0CJaJhhh0CJaJh0h0B*hphhJh0B*CJhphh0B*CJ hphh0B*CJhphh0B*CJhph"  2468:>@b$#x7$8$H$^#`a$gdkigdr2bDln8p & F7$8$H$gdZN 7$8$H$gdZN $7$8$H$a$gdZN & F7$8$H$gdki $7$8$H$a$gdki & F7$8$H$^`gdkin68  *Ƽە惑ynncy\n\n\nc hhh0hh0CJaJhhh0CJaJh0B*hphhah05B*\phh0hX(h0B*CJaJphhh0B*phhpch05B*phh05B*phh_h05B*phh0B*phhwbh0B*phh05B*\phhwbh05B*\ph#  gdr2 $7$8$H$a$gdZN 7$8$H$gdZN \^bdfhjlnrtvxz|~gdr2 $7$8$H$a$gdki*\^bfhlnrtvxz|~㿴 h]h!1h6hhh0CJaJ hhh0hh0CJaJhh0CJaJ hh0h0B*hphh0h"h05B*\phgdr21$gdr2< 001hP:paV/ =!p"#@ $% <001hP:pIX/ =!p"#$% <001hP:p'[/ =!p"#@ $% <001hP:p+/ =!p"#$% =0&P1hP:pN/ =!p"#$% 61hP:pN/ =!p"#$% 61hP:p:Un/ =!p"#$% 21h:p_/ =!p"#$% 21h:pL/ =!p"#$% 61hP:pL/ =!p"#$% DyK yK (http://www.msi.org/$$If!vh555555#v#v:V F470Q!,,55/  4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/  4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ 4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ 4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ 4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,554 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ 4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ 4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ 4 Faf4ytYT$$If!vh555555#v#v:V F40Q!,,55/ /  4 Faf4ytYT$$Ifl!vh55#v#v:V l4 t0+,55/  alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l4 t0+,55~ 5$ /  alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t0,55~ 5$ /  / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / alyt:Un$$Ifl!vh55~ 5$ #v#v~ #v$ :V l t055~ 5$ / /  alyt:UnDd(2l$J  3 @@#" ?$$If!vh55#v#v:V l t06,55/  /  aytr2T$$If!vh55#v#v:V l t0655/  / aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/ aytr2T$$If!vh55#v#v:V l t0655/  /  aytr2TFDd |b  c $A? ?3"`?2/p:WO%Wl0`!d/p:WO%Wۚ@`02xcdd``ed``baV d,FYzP1n:&! KA?H1Z UsC0&dT20 KXB2sSRs"6~.Ob\\i2Usi  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~=>Root Entryn F@+17Data WordDocumentmGObjectPoolp;D[1@+1_1301580600FD[1D[1Ole CompObjfObjInfo  #$%&'*-./012369:;<=>?@ABCDEFILMNQTUX[\]^_`cfghijklmpsvyz{| FMicrosoft Equation 3.0 DS Equation Equation.39q5`| z is i  =1 FMicrosoft Equation 3.0 DS EqEquation Native Q_1301580647E FD[1D[1Ole CompObj fuation Equation.39q5 z is i  =0 FMicrosoft Equation 3.0 DS Equation Equation.39qObjInfo Equation Native  Q_1301580707FD[1D[1Ole  CompObj fObjInfoEquation Native Q_1301580775 FD[1D[15 y it i  =1 FMicrosoft Equation 3.0 DS Equation Equation.39q5 < y it i  =0#LfVJ,1 Ma`(Qy 2|@27 d-܄0sIt'\FL`2*i.{0>gEMf1z8PFO U\X48q(JQ@6+KRsePdk1 ᖄaDd  @ 3 A#"  12zisi=1b%ΝPR)z2nΝPR)zPNG  IHDR@4sRGB pHYsttfxIDATXGWˍ0^r"SG(XJ2pEpK;!O /6F\"x̛g'^>^9y|w`,@;l<22`9 E o)N<(,\}u hPs֟i<@w◘[mƛs͢۩s FsRh2!#k))[@!^h,ο> =_+몶l 12zis=0bv  敫&Rnv  敫&PNG  IHDR<CdsRGB pHYsttfxIIDATHK0ExaXT%@z"8xa wC Q0Θ'C:s.yDY#?yMIѶ?Pb0IFAk#(3H"E GFиt/.aZDFܣLg5Ü~t&-mKp[7gB&+Rɨ=5#mDMC}@6_ul4ll¢{`r Ee]NAÐ`b8W/i5c/Lm].l`#J7ȥ{x>e= ȸ׭'X1&7IENDB`&Dd ( B 3 A#"  12s<sib4=i?¿HHmn4=i?¿HHPNG  IHDR.fKsRGB pHYsttfxIDATHKU 00I CҘh)Htlbl|B߽<QeP*V)CbueuF,r*TsYJXOJXfJ~vf c{ 12yiti=1b[$|97@n/$|9PNG  IHDRABsRGB pHYsttfxIDATXGWn0>$DY2!ҍkߠY>CUx T ?D(~36lH%s>w?,KqݸH=W<&qtBųuXd ܆uDC98e!ȵ8jne6.ERu-/خkϺdF+6zS}N3aOd"OpΪ9&4"zB~ Go]`/)Za(ρ9~Eƫk2`!_; '^ `[$?IENDB`9Dd  D 3 A#"S G 12tibT 8M N?]0רn( 8M N?]PNG  IHDRsRGB pHYsttfxIDAT8Oc?рh tQ`'%/+?DܚaF{̹O0yp CbK@^$.,JP_20F#ft܍q <44ф3 ڃT ]{ GLg$0hW,IENDB`DDd |b  c $A? ?3"`?2c8nJʞj`!bc8nJʞ`(+00xcdd``ed``baV d,FYzP1n:,B@?b X*깡jx|K2B* R vfjv ,L ! ~ Ay ?'rw41 @@ڈq3HA%@W&mJ^Qy L #l&H;!2B͕` 5WLsQ5%@{@&TpAc ơ(Ep`dbR ,.Ieԡ"|f@3X?ŇDd  E 3 A#"  12yit=0 bp:y #q/Tnp:y #q/PNG  IHDR=sRGB pHYsttfxIDATHKV;n0 sY:7Tt39@R؇%ndӲͧ . eJ||"(m }/|i,-OD=9eQم(;0S҃W6C3)9́cU̡e/Y Oz.p:,!.jnʟ{,5zD/+R|yg2&5ٿkl) *8*pA oS n.z,O;ivC^%> ChF)8\BFY)25g\ 5ì[ٞĉ!X=_F 12t<ti bP[D=*H[nP[D=*H[PNG  IHDR,3sRGB pHYsttfx+IDATHK10E'86pip8#17ِh 63 4qt#kfBua##A؋+:dyD4)/6l8Yػɋ(Cmw&ưipA_B{3m Z=,D~4-Yiځ_YD+pQB4_%v{O]֎gAu0#7LY`!)ea=kYEVoRI~":haiU铷 =+9[~KIENDB`HDd hb  c $A? ?3"`?2$@V7߯en `!f$@V7߯e:@4 |4xcdd``ed``baV d,FYzP1n:&B@?b 30< UXRY7&meabM-VK-WMc< si#, 'e, >ľ?m`4bPa.(AW&00,Z0N@e2"We|Of?emK Fk n8#RpeqIj. @ ]` bgat?3JfF Dd A G 3 A #"  12uit=vit+εit bhFw1* 5 UnhFw1* 5 PNG  IHDRmsRGB pHYsttfxIDATXGW;0p6[! ЇS,6tp Dn❱c[I燴hdޛ7ۑZIkR`Ա:uuFnzEE4F /e@.eEפ Wd$Bn&etR->1,֤9Bg8<$(q@% /fŚ%hN|xKŦ K9-d9$m>F}Za~r&G=VdĔg(5rD)  E?~ma<'n,!D1Z)X:F8RG{2`{X1c94~$n[XafhK}nMfx69-yz0`8;v3-ڸtXd [34KZXσf I;~M fjd(@W7`C3MLS!]nW֡n@]QIENDB`FDd  H 3 A!#"U I 12)=exp( it ")1+exp( it "). FMicrosoft Equation 3.0 DS Equation Equation.39qOle 4CompObj(*5fObjInfo+7Equation Native 8WÉ  it = 1 + 2 D 2 + 3 D 3 + 4 D 4 + 5 D 5 + CBI(t) CBIS it + PV(t) RPV it         + 1 SWE it + 2 SWP it + 3 PRSAT it         + 4 INNOV i + 5 SUSINF i + 6 SUSPUR i + 7 GENDER i + 8 AGE i + i ._1303061815.FD[1D[1Ole GCompObj-/HfObjInfo0J FMicrosoft Equation 3.0 DS Equation Equation.39qWï8  CBIS(t) = CBIS  CBIS(t-1) + CBIS .Equation Native K_1303061830'T3FD[1D[1Ole OCompObj24Pf FMicrosoft Equation 3.0 DS Equation Equation.39qWß{  RPV(t) = RPV  RPV(t-1) + RPV . FMicrosoft Equation 3.0 DS EqObjInfo5REquation Native S_13015831918FD[1D[1Ole VA"/>εitb_~o%.aE;n3~o%.aEPNG  IHDRjsRGB pHYsttfxIDAT8O 0 M'ik=:݆Cp&T',/='"B[]y,X9+oO606_ S0{)aTa81¹$LnwAҫ "NC$'&=HQjJJ_CKAamGYDHG6]x--rX^?7q̷/HF68GIENDB`oDd dM I 3 A"#"  12π>0bEQA 8nEQA PNG  IHDR3d6AVsRGB pHYsttfx6IDATHKV;0 DƂC@1x <֑ q&Mv Y ǂm̌y4K[%* _o8Zk1#j02bB*uFcLc`V.IǩH/"u!7:p]907vH52vky˄gCdkBh dRlsnk~">tdRf0YDt2,W1`f졭jt_#ѳX_P[Z:3&RqѱO_8--/G֞ bS&Ѩ/IENDB`[Dd b  c $A? ?3"`?2Wwy=ynȂzPR`!yWwy=ynȂz"@x5GxڥAkA߼6JL )z $B ɆdՃ $ţC(ED<ࡤuΐ V:0<@@1ǧByʲDmPEpVoqT0$#gr^ >i |"pܜwE.{_Ssn2Ÿk[pv=Z\й=g븩ShW1sg| k(ńۃty8޳݃48n$ɲ>k@JŪ}X +h1 YW>B\S\W1o)yYtʳ@wMe=u|=f0O0&c2"a5㐨EگT$8JWZ:-;sB{g]߿Z[VwԞS7ēg';:A|A3gŏGœCۙ͠#z]:/k2[u%X}HASڶu Dd b  c $A? ?3"`?2ԣȯl74-TTU`!ԣȯl74-TT@/@X@Ixڭ_HA3{jyywg|H|Naϓ ^z "zH|z률 @)07;;s(E~hY@;Un?u["ԶmJV_@_#kw^Q Ƌ+Vhh>5lM%>p> !+=xԠ/*pi~0Zfʈ L]Þw nl>Kw0D:3kDua`鄓ÍM9`ax!]+ߵv#ߗPo."Nڿ^&qroEmپv>oty3oq4Į~JCCV.oast 7`aՀwq/'21 Dd Lb  c $A? ?3"`?2VK{`$qP` p2sY`!*K{`$qP` p`.G xڝAhA&iv7mӤ56UAk["$i-b6 TR&@ɭx s(ţ '"͂RE`ܙl`ovgdf $4E[!EjZ3MHedgBFh* ^_G]9d WNJ7KZnQxfy=]Ss,L%SFllXYCG5'cvi%qHn7MRɴH>tjNn$Zi5C6/5a~t[lͦ^l_Ŭ.Y~?C >壂:Pa'8bvRu҅9S5#5`_Gpɍ}a iRNm첼T V[Qrb~j.0QZ>VqD`0l\\ׅ}8}f`а;k}YVj}Y؈[e{|_"]{u짷c '~4b^ {M 2a @yojKO$OQttSbekxoan'Do5P`XeJ"IA8_!MNN_~\/=Dd s J 3 A##"5 ) 12λi~N(0,σ2)bv<,/R(R]nJ<,/R(PNG  IHDR`IsRGB pHYsttfxIDATXGW;N15d#&gEeA&oc{K"g{3_[صJȁ'%ۨeY_YPTf.*a][+V1~QzۅAna!c i$Ǔ%do]ld56RC @K,pAMQ X1P1qJ2ǥ*4kOU&%e.&voc˟Ǘk_7kڷ `31OýCEO|v{խ uXb"_*f PD 7Nԅ,i]+L\6{DwX4m3 ~z!I1bހpra Q1I}2k=F*I&ՎC(V:l% Nձ,s8a1ҤI{hBW%rswV{`^pwED+s4',nH:nR]bZϸ)APU|WkmVPC Ӌ]^*.fTIFlcͨ~=  z`{RCyIWGkJtKΉκ4L>c E/VOބJ"LNPθs+3Z/*]1U=^jZݎIENDB`Dd |b  c $A? ?3"`? 2|i%B3 ô|`!|i%B3 ôT@`0xcdd``^ @c112BYL%bpu c 3X?Dd 8 |b  c $A? ?3"`? 2 qI"R`! qI"D`0zxڝRKQ[u]L4"PBQ@!j56 dS=,ttu gb6q8lŒփv܉HVŹuY9Ay{'wi¯~x} wet+HWW,X~d[ G.Ht4Oɱ3F^:\hD^ W a0#OʕM%\Ui 12ρbYUX 7UOT5n-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0 12ρbYUX 7UOT5՚n-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0 12ρbYUX 7UOT5ʳn-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0 12ρbYUX 7UOT5n-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0 12ρbYUX 7UOT5n-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0 12ρbYUX 7UOT5n-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0 12ρbYUX 7UOT5n-UX 7UOTPNG  IHDR/sRGB pHYsttfxIDAT8Oc?D zz2Bcd?`@1>H\T}8}0c,EW@(e0`!E~ٱ>c,EW@(e&(+\xڭUMhA~v&5&mRC "(X/Dh.$vɛJO^<A*EĮf]] η{ l(ǁY fȘ?c躮ٌ-pgғp \ >{ǀ']PyOle aCompObj<>bfObjInfo?dEquation Native en1-h it ()[] t=s i t i " +z is ln(p is )+(1-z is )ln(1-p is )[] s=1s i " . FMicrosoft Equation 3.0 DS Eq_1301583760BFD[1D[1Ole nCompObjACofObjInfoDquation Equation.39qO4 LL=LL ii=1n " . FMicrosoft Equation 3.0 DS Equation Equation.39qEquation Native rk_1301579546JGFp]1p]1Ole tCompObjFHuf     YZ !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXe\[^]`_abcfghijklmnopqrstuvwxyz{|}~|Z^;$NvQxxHEQE(,ؽy۽\}7w xt[S|1AZmzIIlƧBݹbb_-eץuNze^MA{voO^df /@o>eKyRu ThvoQJ Iz7 LOxY^6:#2nrfZcZ̀ QN77 VXVCzO)Y1i?an8_xK;+1|n?΅Ǧs6fdH?_D 1Q-'7ds?CO:>*>xͿ=4qy?/ŝO.ئqx _)WDd b  c $A? ?3"`? 2)$&^Og4`!$&^Og` +B(+xڭMhAnbIêiJRRQR6RS6Ф,Ƀ'rf E/+͓zW["&ΛYvbnffg @硗$%zZR,`IE u!oQ^8O d1Eq̽N2PƤD!I;Z)C1V BFШ^wX7ד+E%b s[4u%[~е_4~CkVw:nll~=(qu[]7\OUZuq;5t)sl1r1[\g .V ;lp>E kb^S'ͯCȆbJQy6tr) /(qfHԗH>hiW3Nڼ犜r7L$K/r6~xy/OȟYrY h+n_dyU6q}6ٯOi\v)br>d>T}(1s%ȯHSws0rBW-N@lMcX$${foqhL[T"#4UH8`a#lmbנ?ÌW+8`V EWe'Qg1uBloÃADd  R 3 A+#"S G 12sib\|(_ ̒qJ 87n0|(_ ̒qJ PNG  IHDRsRGB pHYsttfxIDAT8Oc?) JÁRF$PzW`@< Txe}`kxp Xs M_*f PyC<­ 6A5IExƎPr`>~0Оsza8@du~), K`F 惹s ȘI˲pIENDB`9Dd  S 3 A,#"S G 12ti bT 8M N?]0#Rn( 8M N?]PNG  IHDRsRGB pHYsttfxIDAT8Oc?рh tQ`'%/+?DܚaF{̹O0yp CbK@^$.,JP_20F#ft܍q <44ф3 ڃT ]{ GLg$0hW,IENDB`tDd b  c $A? ?3"`? 2Cl#on[ˁ\l`!Cl#on[ˁ@e `xڝR=KP=&m,)͂B*b;Tp`Zi+%dWtա?Cpn- W2D ⃛wOsː3!NG0 eVb l\Vf("(p<ax5xa;.0|I &PkznvnkP<[>옼ngnt%?G)aR&VzN?=qĺ(uƅk17nj<"VX~F>'>w>U/s+Lxkc$N޺ΛolXrJ?w~VXrsdWZbCyƈ~z^%\؟H`oʑ{$$If!vh5z 555555#vz #v#v:V F470!,5z 55/  4 Ff4yt T$$If!vh5z 555555#vz #v#v:V F40!,5z 55/  / 4 Ff4yt T$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ 4 Ff4yt T$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ 4 Ff4yt T$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ 4 Ff4yt T$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ 4 Ff4yt T)$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ 4 Ff4p<yt Tkd&t$$IfTF4֞4 Ld|>!z 0!44 Faf4p<yt T)$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ 4 Ff4p<yt TkdMw$$IfTF4֞4 Ld|>!z 0!44 Faf4p<yt T7$$If!vh5z 555555#vz #v#v:V F40!,5z 55/ /  4 Ff4p<yt Tkdtz$$IfTF4֞4 Ld|>!z       0!44 Faf4p<yt T$$IfF!v h5 55555555 5 #v #v :V F470!,,5 5 /  4 FaFf4yt-)kd}$$IfF47 p "6!                    0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 /  4 FaFf4yt-%kdՀ$$IfF4 p "6!          0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kd$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kd$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kdA$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kde$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kd$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kd$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4yt-%kdі$$IfF4 p "6! 0!((((44 FaFf4yt-$$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / 4 FaFf4ytV%kd$$IfF4 p "6! 0!((((44 FaFf4ytV $$IfF!v h5 55555555 5 #v #v :V F40!,,5 5 / /  4 FaFf4ytV%kd$$IfF4 p "6!          0!((((44 FaFf4ytV$$Ifl!vh5565656#v#v6:Vl t06556/  alytD_S$$Ifl!vh5565656#v#v6:Vl t06556/  alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556alytD_S$$Ifl!vh5565656#v#v6:Vl t06556/  alytD_SDd nD  3 @@"?Dd ObjInfoIwEquation Native x_1301577875LFp]1p]1Ole }~2 L=(Pr[t i = i ]) 1-c i .(Pr i=1n " [t i > i ]) c i . FMicrosoft Equation 3.0 DS Equation Equation.39qCompObjKM~fObjInfoNEquation Native C_1303297955QFp]1p]1'P Pr[t i = i ]=h i i  .(1-h i, i -1 ).(1-h i, i -2 )...(1-h i1 ). b  c $A? ?3"`?2O""ŊvWp.Nj+.`!#""ŊvWp.Nj @L"xcdd``nfb``baV d,FYzP1n:&1! KA?H1D>1@0&dT20< 01d++&1Dl\ _˻Dްsi# z0>_ʫʻ2>$'= ]L:`weVy |= 7F=ϊLHFLe 9 ~:o2#?8$.Tg#  _6aPRՄAFFkrByIA#(DL)h0g0aa0`\`պ0a&N0 oD/mFFq"nw&W+ۈGf{%TG^@\121)W2ԡRY`/.Dd | T 3 A-#"  12ci=1$bۣ9{ c3nۣ9{ cPNG  IHDR5q sRGB pHYsttfx*IDATHK10E'06a<6p 9ÁaAMb0#I6(rts! [RxRu M~6;ȁ \zH<& (֚??;XLzŽ)iCQ E>RW}XD Fovߍ Ip/T13-Vݿ/؅+ wݤ:ƿg]Zx_OIENDB`aDd   U 3 A.#"T H 12hit%b{ S^QܮSuWanO S^QܮSuPNG  IHDRZlsRGB pHYsttfxIDAT8O͔ E1d&tHJ2r": e" !\P^1NG@ 5 12hit&b{ S^QܮSuWnO S^QܮSuPNG  IHDRZlsRGB pHYsttfxIDAT8O͔ E1d&tHJ2r": e" !\P^1NG@ 5 12τi'bL{9W#3L(#n {9W#3LPNG  IHDRsRGB pHYsttfxIDAT8Oc?) EÃ_1>0pY %A,|. RՊOo3tIBXGdDI~[t~KkЧC q5^[uB@ߏ{tNNF('8)tHԇC1/~[y!kǑXCa?ѿy#ޠ?@(=akȽ$&,yfQ#"Kq'+*^DO=)߾IŪlPk_},P6kC/=b]@tb=["tґ,qde63@c\lGJ)9Dd p\n0 X 3 A1#"00 $0 12Prti=τi=hiτi⋅1-hi,τi-1⋅1-hi,τi-2⋯(1-hi1))bg{ʘ#hZ(`Cf"n;{ʘ#hZ(`PNG  IHDRM2sRGB pHYsttfxIDATx^]1r:II‡Ǎ(yH'|x+` ;bs>}to[F !2԰d@vgo'gM<6:0ղO\rLfjkpfWh\LIQݐ:#93 5""jGsu×3`a4ys=Z>M@+!`(Y! &6'ؔUezaβul+$Y&gQW=i&IJweo X23 0Z6 k܈rza@tO:;ȑ0 gl .yp) ZI -Zv$Mzj,%B `#Qs2iލ^9+p𠟺8TQj ҹ.szpxWlx|scɏ p02)i!GWǪ`!Qd=gF)& , >B-Z:/Wneo phtC|?6w aV $y@l٘h\K;O׮(o]WZ?"ohuBүI )eavyM hn}-Gp)jÂd˽`0oEOLԷiFj_q$6$s+s)` *DzЊIm4\~ 8` ʈ0\0 L,p藿c CdH oh ^Oi9"Z6 J F\Y Iq<iAol ˆCj]5yRJ>t_ ^Ck Mj=00< >0dpd3wWsk6dL6v٤й馒4euk =2zO/ al$cA#ևIrYH+lB n@_k41|`Bt,0#(RKj00H:񲟨q`$2drL94F|84X-\&0Bf=YG)6!-IO.,cJi$M/)!5bK'$scPnؾ9I&3TP d(IENDB`Dd Hb   c $A ? ?3"`?2_Jitq9`z[`!_Jitq9`z @P&"xڥMhAlIѨUzX=%4i{)[DЃDhZ5 1'c/ 7ɣ DBCAWDYh̛NH o>.-Qƀv&`e"W}WE2* )]Ҙslu8s1dU_Ѻl& gJneyt0Z0hcTӀlVg by;ߍc&8ήG->fky!粿|OWa^U%ig45~Aa.isL_: uN,".^de(f8[L sXלdr^pOȎrUݸux'x-Z}p獏:LveqG:6Mc8M7?n_Ǻ)J/wS={LݿI8.U_Ne[>n݀:O~,B6</K4Ks4a}V9d;ޏ}IyI'Iq] /&~@w=p8%d H%0,\Î84Ӭtx\`0*o{F{5$Dd  b   c $A ? ?3"`?2c\YehNJ"_`!c\YehNJ%S:8xڥMhAnjIMچ&m"FI%9X=DI(Ń$hI*t!&C#^׃^*zYE)ŏ|l3K.,;}ofvx ,F!kil4w^[ZhB`R1q5X\,2!{JU(Tjr#[0tQ"w4( \$b7y |<,OeYݜy_^Y}izw%-q Ed{+ͷ+2u׽GM;߬-T^{]ie'C]au{ٯ^gUI+&mub;]? 7n{FvѯWZhVe/_Ϲ(zΟ o?(9}Pc>~ O{k Y?vQ>"ԣZ?~ցD]#̣.3D-"V[cqs6HBBPf̓92U-`JȧyGf\G$Lv}ن ;'7.DoLz7*oX*aeOD/я~Kl9M9|@v9S[!yM{"yUa{mH?z TM?λ΋}ۯ,@XAŧRԆs $oLg\QFrN̳~8o}8RnH̢?.'mX*tpUaq% _yv.Dd | Y 3 A2#"  12ci=1-bۣ9{ cifnۣ9{ cPNG  IHDR5q sRGB pHYsttfx*IDATHK10E'06a<6p 9ÁaAMb0#I6(rts! [RxRu M~6;ȁ \zH<& (֚??;XLzŽ)iCQ E>RW}XD Fovߍ Ip/T13-Vݿ/؅+ wݤ:ƿg]Zx_OIENDB`/Dd  Z 3 A3#"a U 121-ci=0.b<h脷т~*nh脷т~*PNG  IHDRTW ^_sRGB pHYsttfxIDATXGX;09IV4"+܀4A{h1!(hۉ ?(^k2 yo "DcXScgb\eǢ1zʫG2BsgC`ȧG@뷾i*P]D@O{!`sn7ªvw{\n  ,Σ4] W}`z9#y#(%&Z[Ɏ6eIovHJMtmqߑ3:tTK ꦓ7yS( kK/d $dz4zotrVxa_eʬWT6nnNz}QEl( P*`{ҷ^g6̃EvpY2 SoONIENDB`Dd Hb   c $A ? ?3"`?2I*Allb:Ɲ`!I*Allb:L@!"yxڥkANmҦ"RE)zEzIIM I+]{(_xTMOZOAoxtC !y[|`] |u&"CU4%3\1dԙ WIK̑*!#Uf;b߳៲ߺ1ڢ9^Z\HNz^v`HWw?%$os]iF=+-멜ogCx trRvor! 㑻NuoYSӃ31a==ݽ?{F=5"xC%ujeF0# w]y7BˈۜPzt_uE%#Ơ%nƒ ?Dd  [ 3 A4#"S G 12ci0bZ#!Kf:q?6Sn.#!Kf:q?PNG  IHDRsRGB pHYsttfxIDAT8Oc?) LÃpPzGh@= R`P|0% s@`(`UI^v . Anf<>@2@L+$o=94<v܀r`F5( % `k/>?ڃx4[GYܹɄ@(82f28#GJg IENDB`{Dd tpb   c $A ? ?3"`?2=ەI{6`!=ەI{6 / IgxڥOHQ{v]wݿvuSC QmrZaC]t!b$=tRѡ@:<Ao-! bۼff$ >{?IT} bED&|pԘ$iSGV7By6yTΪ :TDcJ4u>^\]#ƒ]jA%lWGحP;j_'?:$#|r(KӏWFc-w -<yt^laܯq6*7,a28uAڠa$DG_uf],B2"xs0xOB X8-Mau(Ӄ f#a33gp(z J4۠e|Ǽf{Z,]^+,l~La+!^ХI%f~Kg|dfV _|7aTwхM:d7tX`:}CU{ewxǤ}Pt<)A[ wy'us;m^WL.<3dOQ-+?B+޺ښfs,W3lCY/?)zfܢiO_9*Z9r ^2T>}ըfy_if\KpA> "^>ԫ z .8^sTpA/P(G? <׮XL6u N=eg;lpf#J0b]Dth.w×/J&75h w`V;_z씜;Dd b  c $A? ?3"`?2ʡ(,t7a `!Yʡ(,t7&(+'xcdd`` @c112BYL%bpuz!?z!@z!Az!Bz!Cz!Dz!Ez!Fz!Gz!Hz!Iz!Jz!Kz!Lz!Mz!Nz!Oz!Pz!Qz!Rz!Sz!Tz!Uz!Vz!Wz!Xz!Yz ZzFp y\Q#Y+2(88 @GO$V]&elKszAĐi+ˮ'A{M 6 #,D4<cEN8V^bf\fkmt{~pX ūZ(F"_~$()r)1@: CKrTlX Y_gst&   > vq;R'qF{ !"#$%&'()*+c,-l./012Y345F678 9:;<Q=>?#@A B/CDEFWGHIJbKLMNjO~PdQRS4TUVCW X Y Z[\U]^ _ `a b7c$d 1  !"9\qrstw?h./01TpO 8Tg 4j1?N^y 1YAz=\ZQZ!'*X.2@@@ ~WNormalCJ_HaJmH sH tH DA@D Default Paragraph FontRi@R  Table Normal4 l4a (k@(No List>O> |6KEquation !d@"@ $Caption xx5CJ\aJj@j | Table Grid7:V04 @"4 ZOFooter  !.)@1. ZO Page Number4@B4 )RHeader  !>@R> kr Footnote TextCJaJ@&@a@ krFootnote ReferenceH*B'qB (Comment ReferenceCJaJ<< ( Comment TextCJaJ@j@ (Comment Subject5\HH ( Balloon TextCJOJQJ^JaJ>Q@> ~ Body Text 3xCJaJZOZ ?# New Normale^e`5>*OJQJaJnHtHHP@H ?# Body Text 2 dx mHsHtH@O@ ?#Tabelle1CJOJQJmHsH:U@: ?# Hyperlink>*B*^JphB^@B ?# Normal (Web) dd[$\$.W@. ?#Strong 5\^J@O"@ ?# article text"$da$^R@2^ ?#Body Text Indent 2#hdx^h mHsHtH2X@A2 ?#Emphasis 6]^JBORB ?#refs%hd^h`aJ&Oa& ?#email^J0Oq0 ?# yshortcuts^JH+H ?# Endnote Text(CJaJmHsHtHFOF ?# Header CharCJ_HaJmH sH tH 6O6 ?# Style 10 ptCJ^J2O2 ?# medium-font^J@O@ ?#title-link-wrapper^J(O( ?#hidden^J@O@ ?# bibitems1^Jeh@r@BOB 0: My Papers/CJOJPJQJaJTOT /:My Papers Char OJPJQJ_HaJmH sH tH FV@F FollowedHyperlink >*B* phe        !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcd ep$Bg;: !"#$ %hgvqa`^ZUTom4*)+S  &1'2(3)N67890<=> ? @ A B CDEFGHIJKLM*, ep$Bg     !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNpQ#bfmr)>|7oe zzz!z !z!z !z!z!z! z! z!z !z!z!z!z!z!157"8#8$8%8&8'8(8)8*8+8,8-8.8/808182838485868788898:8;8<8=8>8m888888888888888<@GMOSU%U_WW]`&eMejn*noooortxzK}\?'CҒӒɜ5=>ׯ)³M.r{OnX[ 6>hz{u=/d+G_" ha##$$8%9%%%h&i&''''((o)p)7*8***e+f+,m,n,1-2---b.c.////j0k0%1&111s2t22233D4445566i7j7 8 888|9}9 : :::;;F<G<<=B>C>>>????@@#A$ABBBB"C#CCC,D-DDDEEHFIFFFkGlGEHFHHHIIJJkJlJKKKKLLLLpMqMNNOOhOiOPPPP+Q,Q!RRuSvSSSTTjUkU7V8V/W0WXXXXNYOY|Z}ZA[B[[[\\]]]]D^E^^^^^___ _"_$_&_(_*_+_4_8_<_@_D_H_I_^_b_f_j_n_r_s________________________`` `````"`'`,`1`2`A`E`I`N`R`W`X`n`r`v`z`~````b b0b1bl?l@lAlBlClDlElFlGlHlIlJlKlLlMlmmnmmmmmn^pupNuduEzdz~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~hijklmnop}ډVJܗl`Pu=UZy9:kEFMT#'5-lWCD \  -.8;?BFIMPTW[dilm{"',16;@EJOTYZx}      , - . / 0 1 6 : ; Y Z [ \ ] ^ c h i             !!!!!!!!!!!!!!!!!!!!!!!!!!"""""""" "!"""9"="A"B"C"D"E"F"G"H"I"`"d"h"l"m"n"o"p"q"r"s""""""""""""""""""""""""## ########>#B#F#J#N#R#V#Z#[#\#]########################$$($O$P$d$l$t$|$}$$$$$$$$$$$$$$$$$$$$%%%%%%$%+%5%<%C%J%K%N%X%_%i%p%v%~%%%%%%%%%%%%%%%%&&&$&)&0&1&P&Q&Y&`&h&o&p&&&&&&&&&&&&&&&&&&&' ' '''''$'+','9':';'D'K'L'e'f'g'p'w'x''''''''''''''''(((( (0(8(?(G(N(V(](^(n(o(p(q(r(w(~((((((((((((((((((((')()3)R)S)r)})$*%*++++,,]-^- . .{.|.L/M/ 0 0}0~0001K2L2132333{4|4;5<5556677X8Y888091999::5;6;<<<<==>>i?j?K@L@@@3A4A B BBB CCC'D(DDDMENE'F(FFFGGGG}H~HII.J/JJJ>K?KKKNLOLMMMM+N,NNNOOPPQQbRcRRRSSrTTTTUUVVW7W[W{WWWXlXXXXXXXXXXXXXXXXXXXXXXXXXXXXYYYYYYYYY Y Y Y YY,Y-YXYYY Z&Z)Z*Z+Z,Z>Z?ZAZBZCZDZEZFZGZHZIZJZKZLZMZNZOZPZQZRZSZTZt[[[J\\\]]]?]]]^v^^__:___5`````a'a~aabWbbbbb0cIcc d|dd3e4e5eUeefDfwfffDggggggggggggghphhhNiiijmjj kkkklxlllloooo pp/pGpSp`plpypppppp q qqqqqqqqqqqqRqSqeqqqqqqqqqqqqqqqq'H`Ɂʁˁށ߁'()*;<=>?@ABCEOPQSVXYZ\^_`bdefghjlmnopqrtvwxyz{|~˂̂͂ނ߂ ",-.03D_vxʃ܃Ua˄  !#BDEFGHJLMNOPQRTVWX_abc!!!!!!!!!!!!!!!!!!!!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!Y!O !4 !!!!!!!!!!!!!!{5!UE!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!!:!U!;-@!U!!:!!:!{5!UE!o%!Y!Y!Y!Y!Y!Y!Y!!!!!!!!!!!!!!!!!!!!!!!!!!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!Y!;-@!;-@!'Op!Z!!0!;-@!id!Y!o%!Y!C`!{5!UE!Y!U!!:!Y!id!!!Y!{5! !UE!!*!{5!{5!!!Y!*!e!{5!o%!!!Y! j!{5!!!Y!oJ!{5!o%!!!!0!!!!:!{5!!!oJ!!!Y!oJ!Y!Y!!:!UE!Y!!0!!*!Y!;-@!Y!U!Y!!!!!!!Y!!!!!!!{5!Y!!:!Y!!:!!{5!o%!Y! !Y!o%!{5! !{5!Z!{5!Y!!{5![{!!:!!0!8P!Y![{!Y!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!_!_!_!_!_!_!_!_!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!GV!GV115111111115!!!!!^Fv:v:ōv:Lv:ōFv:v:Lv:dFv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:u:Fv:v:Lv:d!j!!!!!!!!O!Y!^!Y!Y!Y!Y!Y!Y!!!^$ssR:$ss)$sss$sss$sss$sss$ss$ss$sss$sss$sss$sss$sss$ss$ss$sss$sss$sss$sss$ss$sss$sss$ss!s!s!s!s!s!s!s!s!s!s!s!!!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!v:!! "!!!!Y!o%!Y!8P!Y!8P!Y!oJ!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!Y!U!!:!Au! !!Y!UE!UE!vZ!;-@!Y!!:!Y!Y!o%!{5!!!oJ!U!!!Y!*!oJ!!!Y!Y!{5!oJ!Z!Y!!:! !o%!8P!Y!o%!w)!!EC !4 !EC !Y!U !!id!![!![!o%! !Y! !4 !EC !Y!*!Y!*!Y!!0!o%!Y!{5!Y!U! !!:!Y!oJ!!0!*!Y! !{5!!0!!!GV!! ssssssssssssssc sssssssssssss sssssssssssss sssssssssssss sssssssssssss sssssssssssss ssssssss ssssssss ssssssss!!!!!!!Z v:v:v:v:v:v:v:v:v:v:ōZ v:v:v:v:v:v:v:v:v:v:dZ v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:u:Z v:v:v:v:v:v:v:v:v:v:d!!!!!!Fs^s^s^ssFs^s^s^s^s^s^sFs^s^s^s^s^s^sFs^s^s^s^s^s^sFs^s^s^s^s^s^sFs^s^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^sFs^s^s^s^s^s      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~Fs^s^s^s^sFs^s^s^s^sFs^s^s^s^sFs^s^s^s^sFs^s^s^s^s^s^sFs^s^s^s^s^s^sFs^s^s^ssFs^s^s^ssFs^s^s^ssFs^s^s^ssFs^s^s^ss!!!!!!!<!Y!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!O !EC !O !![!Y!% !Y!l !]3 !Y!% !Y!]3 !Y! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!v:      v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v: v:         Ec       2      9         A         V                   @3                8! !!v:!!!!!!!!!__^^ss^s^^ss^  s s s s h jh jh ;h jh jh GVh jh jh GVh jh jh GVh jh jh #h jh jh jh #h jh jh jh h h h ^h h h ^h sh sj^ss^s^s^ss^^^s^^^s^^ssh jh jh ;h jh jh GVh jh jh GVh jh jh GVh jh jh #h jh jh jh #h jh jh jh h h h ^h h h ^h sh sj^ss    I I I I HՃHHHv:o ^^ss^^^s ^ss^^^s^s^s7J000+233:j?IJP@0@0@0@0@0@0@0@00 12  !"9\qrstw?h./01TdopO 8Tg 4j1?N^y 1YAz=\Z)QZ!'*X.157"8#8$8%8&8'8(8)8*8+8,8-8.8/808182838485868788898:8;8<8=8>8m888888888888888<@GMOSU%U_WW]`&eMejn*noooortxzK}\?'CҒӒɜ5=>ׯ)³M.r{OnX[ 6>hz{u=/d+G_" haJ##$$8%9%%%h&i&''''((o)p)7*8***e+f+,m,n,1-2---b.c.////j0k0%1&111s2t22233D4445566i7j7 8 888|9}9 : :::;;F<G<<=B>C>>>????@@#A$ABBBB"C#CCC,D-DDDEEHFIFFFkGlGEHFHHHIIJJkJlJKKKKLLLLpMqMNNOOhOiOPPPP+Q,Q!RRuSvSSSTTjUkU7V8V/W0WXXXXNYOY|Z}ZA[B[[[\\]]]]D^E^^^^^___ _"_$_&_(_*_+_4_8_<_@_D_H_I_^_b_f_j_n_r_s________________________`` `````"`'`,`1`2`A`E`I`N`R`W`X`n`r`v`z`~`````bb b0b1bl?l@lAlBlClDlElFlGlHlIlJlKlLlMlmmnmmmmmn^pupNuduEzdz~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~hijklmnop}ډVJܗl`Pu=UZy9:kEFMT#'5-lWCD \  -.8;?BFIMPTW[dilm{"',16;@EJOTYZx}      , - . / 0 1 6 : ; Y Z [ \ ] ^ c h i             !!!!!!!!!!!!!!!!!!!!!!!!!!"""""""" "!"""9"="A"B"C"D"E"F"G"H"I"`"d"h"l"m"n"o"p"q"r"s""""""""""""""""""""""""## ########>#B#F#J#N#R#V#Z#[#\#]########################$$($O$P$d$l$t$|$}$$$$$$$$$$$$$$$$$$$$%%%%%%$%+%5%<%C%J%K%N%X%_%i%p%v%~%%%%%%%%%%%%%%%%&&&$&)&0&1&P&Q&Y&`&h&o&p&&&&&&&&&&&&&&&&&&&' ' '''''$'+','9':';'D'K'L'e'f'g'p'w'x''''''''''''''''(((( (0(8(?(G(N(V(](^(n(o(p(q(r(w(~((((((((((((((((((((')()3)R)S)p)q)r)})$*%*++++,,]-^- . .{.|.L/M/ 0 0}0~0001K2L2132333{4|4;5<5556677X8Y888091999::5;6;<<<<==>>i?j?K@L@@@3A4A B BBB CCC'D(DDDMENE'F(FFFGGGG}H~HII.J/JJJ>K?KKKNLOLMMMM+N,NNNOOPPQQbRcRRRSSrTTTTUUVVW7W[W{WWWXlXXXXXXXXXXXXXXXXXXXXXXXXXXXXYYYYYYYYY Y Y Y YY,Y-YXYYY Z&Z)Z*Z+Z,Z>Z?ZAZBZCZDZEZFZGZHZIZJZKZLZMZNZOZPZQZRZSZTZt[[[J\\\]]]?]]]^v^^__:___5`````a'a~aabWbbbbb0cIcc d|dd3e4e5eUeefDfwfffDggggggggggggghphhhNiiijmjj kkkklxlllloooooooooooooooooooooop p p p pppp#p$p-p.p/p;p?@ABCDEOPQRSUVXYZ[\^_`abdefghijlmnopqrstvwxyz{|}~˂̂͂ނ߂ !",-./023D_vwxɃʃ܃Ua˄   !"#ABDEFGHIJLMNOPQRSTVWXYZ[\]^_abcdefghijklmnopqrstuvwxyz{|}~00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 0000000 0 0 0 00000000000000000000000000000000#0#0#0#0#0#0#0#0#0#0#00000000000000000%%000"0"00000 0 0 0 0 0 0 0 00"0"0"0"0"0"0"0"0"0"000"0"0"0"0"0"0"0"00000000000000000000000000000"0"00000"0"0"0"000000000"0"00000000000"0"0"0"0"0"000000"000"0"000"0"0"0"0"0000"0"0"0"0"0"0"0"0000000000000000000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00000000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00000000000000000000000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000000000000000000000000000000000000000H00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000/0/0/0/0/0/0/0/0/0/0/0/0/0/0/0/0/0000000000000000000000000 000 00 00 00 00 00 000 0 0 00 00 00 00 00 00 0 0 00 00 00 00 00 00 0 0 00 00 00 00 00 00 0 0 00 00 00 00 00 00 0 0 00 00 00 00 00 00 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 00 0 0 0 0 0 0 0 00 0 00000 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0000000 0 0 0 0 0 00 00 00 0 0 00 00 00 0 0 00 00 00 0 0 00 00 00 0 0 00 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 00 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 00 00 00 0 0 00 00 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 00000000000#0#0#0"0"0000000000000%00000000000 0 0 0 0 000"0"0"0"0"0"0"0"0"00000"0"0000000#0000000000"0"000000000000"0"0000000"0"0#0#00000000000000000000000"0"000000000/0/0/0/0/0/0/0/0/0/0/0/0/0/0/0/0/000000000000000000000000000000000000000000x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x0x 0| 0000000 0 000000 0 000000000 0 000000000 0 0000000000000000000 0 00 00000 000000000 0 000000000000000000000000100000010000001000000000000 000000 0000000000000000000000000000000000000000 0 0000@0@0@0@000000000000000000000 000 00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 0 0 000 0 0 0 000 0 0 000 0 00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 12  !"9\qrstw?h./01opO 8Tg 4j1?N^y 1YZQZ!'*X.157"8#8$8%8&8'8(8)8*8+8,8-8.8/808182838485868788898:8;8<8=8>8m8888888888888<@GMOSU%U_WW]&eMejn*noooortxzK}\?'CҒӒɜ5>ׯ)³M.r{OnX[ 6>hz{u/+G_" haJ##$$8%9%%%h&i&''''((o)p)7*8***e+f+,m,n,1-2---b.c.////j0k0%1&111s2t22233D4445566i7j7 888|9}9 : :::;;F<G<<=B>C>>>????@@#A$ABBB"C#CCC,D-DDDEHFIFFFkGlGEHFHHHIIJJkJlJKKKLLLLqMNNOOhOiOPPPP+Q,Q!RRuSvSSSTTjUkU8V/WXXXXNYOY|Z}ZA[B[[\\]]D^E^^^___ _"_$_&_(_*_4_8_<_@_D_H_^_b_f_j_n_r_____________________`` ````"`'`,`1`A`E`I`N`R`W`n`r`v`z`~````bb b0b1b#B#N###############($O$P$d$|$}$g'w'x'''''''''(3)R)S)p)q)r)+++,,]-^- . .{.|.L/M/ 0 0~0001L2132333{4|4;5<567Y8199:=>>i?j?K@L@B CC'DNE'F(FFGGGG}H~HIIJOLMM,NNNOOPPQcRUUVVW7W[W{WWWXlXXXXYXYYMZNZOZPZQZRZSZTZt[[J\\\]]?]]^v^^__:___5`````'a~aabWbbbbIcc d|dd4e5eUeeffDggggghphhNiiijkkklxlllloooppSqeqqq ~000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000@0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 0000000 0 0 0 0000000000000000000000000000#0#0#0#0#0#0#0#0#0#0#00000000000000000%0%0000"0"00000 0 0 0 0 0 0 0 00"0"0"0"0"0"0"0"0"000"0"0"0"0"0"0"0"0000000000000000000000000000"00000"0"0"0"000000000"0"000000000"0"0"0"0"0"000000"000"0"000"0"0"0"000"0"0"0"0"0"0"0"0000000000000@0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 00000000 0 0  0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0000000000000000000000Z00aZ00aZ00$bZ00\bZ00 bZ0 0 Z0 0 cZ00PcZ00Z004Z00%Z00%Z00䞴Z00Z00TZ00̣Z0 0!Z0"0#<Z0$0%tZ0607 Z0809X Z02Z02Z0:0Z02@0Z02d4]Z02cZ02a@0@0@0@0@0@00000000000000000000000000000000000000000000000000000000000000Z03C00000000000Z03<0000000@/0/0/0/0/0/0/0/0/0/0/0/0/0/0/0/00Z03 0000Z03@0@ 000000@0@0@0@0Z04 DZ04  Z0!4 @0@0@0@0@0@0@0@000@0Z04@0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 @0 Z04Z04Z04Z04@0  @0 Z00@D@0 @1 Z00@0 @ Z0 0 Z04Ї@5|/ @0 @5|/ Z0 0X/Z04\Z04Z04Z04Z0\4 Z0\4Z0\4Z0\4Z0\4Z0a4 Z0a4b$Z0\4 Z0\4]Z0\4 0 Z0\4 Z0\4 Z0\4 Z0\4 0Z0d4 Z0d4ex0Z0g4h̒000000@0X0300[h00h00@0@0@0@0@0X03@0@0@0@0Z04XZ04WZ04U@0X03,wX03~@0@0@000=00=00=00=0}0#J0a70}0$00E@0@000(Z030Z0V4FW 00@0@00000-@0@0Z000@0@"000 X03iXzX03hX03f@0@0@0@0@0@0@0@00000 Z0Q4RZ0Q4@0@0@0@0@0@0@0@/0@/0@/0@/0@/0@/0@/0@/0@/0@/0@/0@/0@/0@/0@/0@0Z084@0Z03B|X03AX03?@0@0@0@0@0@0@0@0Z03?Z0n4zZ038}@0Z0F4@0@0@0@0@0 @0 Z0N4@0@ 0@ 0Z0y4zZ00f @0@0@0@0@0@0@0Z04 @0 @0@0Z0 0Z0 0Z0|4Z0 0Z04 Z0 0 @0@0Z04Z04Z04@0Z04sZ04 Z0 0 Z04sZ04~Z04~Z04|Z04sZ04zZ04tZ04tZ04sZ04q@0Z04z Z04s 0@0Z00`c=0Z00s=0@000@ 0@ 0@0Z00@ 0@0@0@0Z0 0 0] ''''''))))))555@BMZZZZZerrrrrrrrrrr}      !"#$%&'()*+,-./01254679 W1ZP,?QUvǫ3{xRRm(=,HYbOm vmwhyO}}͂z HmIe>88F@i%   7f "0A GHwIJLLMNN}O&PdQR.TV_fakEsy_,4zh n*GMOPRUWXZ]^`bcefhijkmnpsvx{} "',.02468:;<>?@ADFvy|~ YZX65@D|ɤ:A=H$OV\bgHnuBvLv`vzvvvvvvvww*w0wIwYwowwwyiyyyy zz+z?z`zmzzzzz{'{F{a{s{{{{{{|3|O|k||N}}}}~*~E~~~L2ƀހ&q  \er;) 8%FGGVGGGGG4HSHHHHIiIJJ?KKLLWMMMM%NYNNN OJOOOO PLPPPhQQpRRR S>SrSMTZ_ahoss{AڈzHB ȵjbHJKLNQSTVY[\_adgloqrtuwyz|~     !#$%&()*+-/13579=BCEGpqrstuwxz{}I^I~II7fOfRf  *+ !",-`jk}BVXYcdr|} _ikXbd9CE!+,xS)k)n)UUUUUUUUUEVOVTVVVVVVVVVVW3W5WcWwWyWWWWWWWWWWWXX=XGXHXsXXXXXXX_:#:##:##:##:###:::#::#######::##:_:####:#:::##:#::  $)025<BIMTWZaelory}!!!!!!!!!!!!!!  aQQQ8@F(]'(  F H'_( k3  s"*?` l c $X99?H'_( m 6(QwQwmC"?'y (`" n c $o$' o 6'QwQwoC"?$K' '`" p c $o' q 6QwQwqC"? v `" r c $o.0ZB s S D<TB t C Db c ZB u S D  n >   +* Q3  s"*?h R S X?" +* S <-R)R)S"^ - T 6&R)R)T"$<$* & U 6%R)R)U" lP %h" V S o"$x*h" W S o"h" X S o" h" Y S o" " Z <R)R)Z" ," PB [  "'d 'VB \ # )"Ha\" ] 3 o"|' ^ <R)R)^"1'\ \" _ 3 o"|(# ` <R)R)`"~'#  a <R)R)a"M'2 PB b  "PB c  " pT6!PB d  "!!PB e  "dET!ZB f S D@ (#D ' g C gC"?8 `   h C  hC"?`   TB ) c $DjJn v c $v̙"? P2 w 3 ̙"?P2 x 3 ̙"?P2 y 3 ̙"?P2 z 3 ̙"?P2 { 3 ̙"?P2 | 3 ̙"?P2 } 3 ̙"?P2 ~ 3 ̙"?P2  3 ̙"? P2  3 ̙"? P2  3 ̙"? P2  3 ̙"? P2  3 ̙"? P2  3 ̙"?P2  3 ̙"?DB   "?n  c $̙"? n  c $̙"? zb dH-  #" ?z  c $)̙"?dZ )z  c $!̙"?Wv !z  c $ ̙"?M W!  z  c $̙"?"db$ z  c $̙"?$d~& z  c $̙"?U'<) z  c $̙"?)di+ z  c $̙"?+dH- n  c $̙"? P2  3 ̙"?)P2  3 ̙"?P2  3 ̙"?(P2  3 ̙"?'P2  3 ̙"?&P2  3 ̙"?%P2  3 ̙"?$P2  3 ̙"?#P2  3 ̙"?"P2  3 ̙"?!P2  3 ̙"? P2  3 ̙"?P2  3 ̙"?P2  3 ̙"?P2  3 ̙"?DB   "?n  c $#̙"? #n  c $"̙"?* "zb dH-  #" ?z  c $̙"?dZ z  c $̙"?Wv z  c $/̙"?M W! /z  c $,̙"?"db$ ,z  c $.̙"?$d~& .z  c $6̙"?U'<) 6z  c $0̙"?)di+ 0z  c $$̙"?+dH- $B S  ?6fPfl l%l&l'l(l)l*l+l,l-l.l/l0l1l2l3l4l5l8ll)&Z'Z,Z-Z.Z/Z0Z1Z2Z3Z4Z5Z6Z7Z8Z9Z:Z;Z8m888U%U7d?d&eMen*nooOX6>h  (a##NNQ!R^^_*_+_H_I_r_s_______``1`2`W`X``bObRbccme~eeef\fif}f5gxgggg0h?hohhhhiiiiniii jjUjZjj k@kEkkklMlmmp}3kFMW&U:;ABHIOPVWcdlmSTYZ  5 6 : ; b c h i        !!!"""A"I"l"s"""""##Z#]####$($$''(()3)r)})>>UUVVW7W[WWWXlXXXX Y,Y ZZTZZZ\\``a'abb0cIcshhijloppppppqqqTqqqqqqz|%'^`*CEQSTVZ\`bhjrt|~ ".013vxȃʃU !#@BHJRTX_c}~?@xy;<Z[LN``,,^-a-M/P/33|44GG__gglloooop ppp"p$p,p/p:p|5Q3X bkn{S|e :P%/,jh>24J.ƾW=V f@h!wDԻ #(Fh`UQ$NoS+rWFX8[P6W45_BExiLim4@VgoLL2WVW lluvStbNM<AU^ JGOiu*35I{LM !8` [x(,1RV'X6{DL#i*,_hl7&6HQ   > > > D #e _m p (  H' yT 3h n Dp R( + W d o ]p s N" (. `; PN /r Q" ]I 'i :i 3CWd:!)-6cgO.%&6AW}`xj,=*dZxo/'0<YUbdt0-08:SAk~;jo7x= A=lp[] G% )50;p>,A(J;o=tw:} Y+QV] 1:MPdr/XZuA.Sditw#.dHJ9UUis'7[B`EALTM\fxxG-0&5CCQSZOeps)q 6 I5 HE F Y !P!!M!!! 2!9>!M!M\!\!k!"R#"*"+"5"@"g"|""#P#Nh#i#p#$$y$D$G$xO$|h$T|$%l%)%<%o%:%&&-&;B&yN&_|&`~&{'' '!'0'6':'I'J's'/(5(KA(Y(Y(lc(u(p)!)%)K)).)3)C)=F)kW)Nm)|)**o*?**++*!+%+++=+J+|+~++ ,V,d!,'*,Q,R,T,n,\p,r -(-8-x8-@- U-Z- .)H.I.K.R.~W.c/Fy/0r"0#050I0j011'1(1;81 E1]2[ 2J'232K2r2x23(303U3!_3!4 4*'4p=4H4`_47j4l4s4 }4d5.5=5O5R5Sd5h56!6P/6o0634656B6C6r6x6Ly6@77-757V7hn7`u78_.8F28C889G8^K8`8j8k99!9*909 4979D9YJ9tZ9g9j9Gj9j9 :$:@:j`:^q:7;';*;%?;A;W;[;E];i;t;3<uG<R<S<]<7a<+l<D=i"=g[=g>>A>]>Td>Nn>}>>'(?,?W9?9?_d?p?1@AR@S@ W@e@'x@H"A ?A_AgA* B&Bb#B#B'B{1B9BF,UF[FfFjFwF|FGG GGN G0G8GeG_HH=H}kHEpHqHI;I II_)IVI\IkI$vI~IJJ !JCJ0-Kv/K|6K6K7KfKuKzKLLtYLMMw0MJMaMclM}MNSN0mNtNNOKOAOHODROUOgOkOlOvOPP@%PX_PKuP8xP{PFQQO Q#Q'Q/+QoPQmQ%RPRiRSSBSYSD_SzwSO!T)T]+T%4T@T~PT^TrTvTvzT=U9UYUzhUTnUq|U2~UVfIVaViVI}Vv}VW5W3RWL*X0Xc0X?XaDXIX.UXwXN2Y@YNY=SY-ZY[YgYliYvY_ZoZZ=Z]DZbZjZI[![G'['[([y6[cD[ZK[R[_[zq[ r[ \M1\2\=\@K\P\e\d]#]*%]E8]E]^ 1^C9^D^T^lY^r^__<'_FP_P_T_G]_ `3`\`b`g`-l`}&aLa#[a`eabnb7bZbreblbrvbxbc!c$c`,cp6c&Zcgcd#d|$d#)dIdpidldrd} e ee3ee%e%{yz!zU!z(zSXhikbxL0=@JBKUy\t`e{=5ABlU1Z}rBI(Z} g<+Ng~ j%*{9k?]_Tf[g=| I,RY_w*z8A)E7utf'45Tw&;TZP-?#T0w57wH [*2Y.{* ,jv)*+6'<a=_?b7StYb#ijEltwT ,6GvVW]e8ifp 7'(12::fGk$%7BfB!@DH| e!d9^huV~q )-L$l V/1cG,Wj M.5tLOQVk#@9NQTcAfmj?!$1P=@{}c))5M?cX24]l vF%)-6\Be`mWu}-8BCELfl$(23Jlmzq#=E.}2D5<mI eAzERKT{-Y3 B]Eg ws " ->O[v"$1BbKwKK_f`g)Li2A\\c(euj9Iqdfflrkq#*hefhHv, W[eiHt d8zQ (;=nCZOXhnx ,N}Xpc4/%.$BU  DQ10%1gva$(F@OW]%h1.@'FROcX5d;W` I!B8CD%s( ::OZ7l0 Jd9TJdGma=OUu48`?IQla\b%hMq \+;CM^(Zcj \!,!.L5YAZ]:"L%/w99@%F|Sck zIz,2/IIQXjn "S]jt4 W5\\h]i9kEz|!}bbcmo v/44>C+Uii1EHQRY[zi<OjsvB!3t77Bg{d,&())UQTZcKi |M|{$2.a^y1VW[b*q'SQ_U|jV1:@Bh<N+VC.ZsXW3LOWUULQRy {$)OQqw,=avnzkra{ G7=L3Oeceo"pD3TlL9Zs ;14<S ]hc ZGI7PZak"w|h}6CaV*o:.~8W}e t!3)}6]C LgMChvz Zmm{/:CUV jn ^qw O}cXo fN"U#M,PUg4jpn!"&01nOsT # *xRU_Tetlp{:  (M5L]m&;{?OjRt]!;DR^nsw[yW,07s@[z>"%}-.176PS~W8gs=t+Gbh]yz +99KJLbNbYz|$2|Sdffszyy~O)EH%*?@2On\x6~`,5'77ZNZ\O}q} *K[{ /wWX3cs_yFVPtbi 5H=S?EtUg[  eA$M]acdu&APHKR[`xz#cQZ!\Pd|e<j@n"~k"]#*"=X'y'(*a-!1F/RX ,0>DJBQ[VDKNLToX[+DQFW[T_`) ~%q. 0GL+37"8/LTht\?^_ _"_$_&_(_*_+_4_8_<_@_D_H_I_^_b_f_j_n_r_s________________________`` `````"`'`,`1`2`A`E`I`N`R`W`X`n`r`v`z`~```1b#B#F#J#N#R#V#Z#[#\#]#######################P$d$l$t$|$}$$$$$$$$$$$$$%%%%+%<%J%K%N%_%p%~%%%%%%%%%%%%&&$&0&1&P&Q&`&o&p&&&&&&&&&&&&&& ' ''''+','9':';'K'L'e'f'g'w'x'''''''''''''(( (0(?(N(](^(n(o(p(q(r(w(~(((((((((((((((((((()U Y-Y]]__``bb4e5eggll1J 0Iw01"EN.InstantFormatN~ EN.Layout~ EN.Librariesܜ~w<ENInstantFormat><Enabled>1</Enabled><ScanUnformatted>1</ScanUnformatted><ScanChanges>1</ScanChanges></ENInstantFormat>`<ENLayout><Style>J Marketing Research</Style><LeftDelim>{</LeftDelim><RightDelim>}</RightDelim><FontName>Times New Roman</FontName><FontSize>12</FontSize><ReflistTitle></ReflistTitle><StartingRefnum>1</StartingRefnum><FirstLineIndent>0</FirstLineIndent><HangingIndent>720</HangingIndent><LineSpacing>0</LineSpacing><SpaceAfter>0</SpaceAfter></ENLayout>L<ENLibraries><Libraries><item>Citations.enl</item></Libraries></ENLibraries>@KK8T KKPprswy%%%%'%(%)TUVWYZ]P@PPPP@PP@PPP@PPPP@PP@PP@PPP@PP,@PPP@P P PPPP(@PP8@PP@@P"P$PL@PPPRP@PP@PP @PP@PUnknownDISSERTATIONSLAM Normal.dotslam3Microsoft Office Word@^в@2@D@R&e\3՜.+,D՜.+,8 hp|   f=k'  DISSERTATION Title 8@ _PID_HLINKSA`'http://www.msi.org/  FMicrosoft Office Word Document MSWordDocWord.Document.89q Gz Times New Roman5Symbol3& z Arial5"System5& zaTahomaABook Antiqua[ STXinweiArial Unicode MS?5 z Courier New;Wingdings"qh&ft&\3e=f\3e=f!p~4dkk 3qHX ?Pd2 DISSERTATIONSLAMslaml                   CompObjq