Influence of inconsistent brand positioning on brand attitude & attitudinal loyalty

(1)

Influence of inconsistent brand positioning

on brand attitude & attitudinal loyalty

“Determination of the relationship between inconsistent brand positioning, brand attitude and

attitudinal loyalty, when considering the moderating effects of functional congruity and self-consistency for consumer goods.”

By

Ilse Zwaan

University of Groningen

Faculty of Economic and Business, MSC Marketing

Qualification: “Master thesis”

June, 2013

Supervisor: Dr. K.J. Alsem

Second supervisor: Dr. J. Liu

(2)

Master thesis Ilse Zwaan 2

PREFACE

In my opinion, branding is attaining a larger importance nowadays for companies to differentiate themselves from their competitors. Furthermore, I think branding plays an important role in our society nowadays. Due to this crucial role of branding for both

companies and society, I found branding to be a very interesting research topic for my thesis. Therefore, last January when I heard that I was selected to participate in the Branding Master thesis group I was really content and immediately started writing on my thesis with great enthusiasm.

I can conclude that last few months, while writing this thesis, were a true learning experience for me. I experienced that writing a thesis contains several moments of ups and downs. For example a successful moment, when finding the exact piece of relevant literature. But a disappointing moment when finding out, that after all one of the proposed constructs was not measurable. Luckily, I also learned that there was always a solution for these encountered problems. Writing this thesis would not have been possible without the help and support of my supervisor Dr. K.J. Alsem and fellow group-members of the Branding Master thesis group. Dr. K.J. Alsem helped me when this was needful throughout the process of writing this thesis by providing me with some guidelines and feedback. Furthermore, my fellow group-members helped me by discussing difficulties of writing a thesis together. I would like to thank Dr. K.J. Alsem and all my fellow group-members for all their help and support.

Overall I am really satisfied with this study and I hope all readers will enjoy reading my thesis.

Ilse Zwaan

(3)

MANAGEMENT SUMMARY

Many academic researchers claim that brands should be consistent in their positioning, extensions and communication (E.g. Keller, 1998; Buchanan, Simmons & Bickart, 1999). However, others state that brand consistency might not be that important after all (E.g. Charters, 2009; Kay, 2005). This study examines this discrepancy in literature, by focussing on the effects of inconsistent brand positioning, with reference to showing different target groups in advertisements, on attitudinal loyalty and brand attitude. It was expected that inconsistent brand positioning would have a negative effect on attitudinal loyalty and brand attitudes since in general academic researchers claimed that brand consistency leads to higher brand equity, of which loyalty and therefore also brand attitudes are considered to be a part of (Buchanan et al., 1999; Keller, 1993). Next, it was found in literature that brands are

evaluated by consumers on both symbolic- and functional attributes (E.g. Bhat & Reddy, 1998). These evaluations are captured by the concepts, self-consistency and functional congruity (Kressman Sirgy, Herrmann, Huber, Huber & Lee, 2006). It was expected that the proposed main relationships of this study were moderated by these two concepts, since they were found to have a direct effect on brand loyalty, of which attitudinal loyalty and therefore brand attitudes are considered to be a part of (Kressman, et al., 2006; Oliver, 1999; Keller, 2008). For self-consistency a strengthening effect was expected since the form of brand inconsistency, tested in this study, can only directly influence the symbolic evaluation of a brand. Furthermore, for functional congruity a weakening effect was expected. The research question of this study is;

“How does inconsistent brand positioning, with reference to the target group shown in brand ads, influence brand attitude and attitudinal loyalty and how does the level of

functional congruity and self-consistency influence this relationship?”

(4)

brand and therefore customers could even like inconsistencies in branding (Charters, 2009). For both expected moderators, no significant effects were found. The lack of effects could be explained by theory of Oliver (1999). Oliver (1999) was interpreted as “When there is a state of ultimately loyalty, situational factors have no influence on the bond of loyalty anymore.”. Considering this line of reasoning, it could be the case that both moderating variables, as situational factors, had no significant effect since in an ultimate state of loyalty, the concept is not influenced by situational factors at all.

By examining specific effects of a specified form of brand inconsistency, this study solved a small part of the existing gap in literature with respect to brand consistency. Next to this theoretical implication, this study showed that brand managers, involved with comparable brands, products and target groups as the research stimuli of this study, can use both

inconsistent- and consistent brand positioning, with respect to the target group shown in their advertisements, without significantly affecting attitudinal loyalty and brand attitudes of consumers. Furthermore, brand managers should be aware of the fact that self-consistency and functional congruity of their customers, do not influence the relationship between the consistency level of brand positioning, attitudinal loyalty and brand attitudes. Future research should be conducted on the same research topic, only with other types of brands, products, target groups, time paths, level of exposures, moderating- and dependent variables, like customer involvement and brand awareness.

Key words: Inconsistent brand positioning, attitudinal loyalty, brand attitude,

self-consistency and functional congruity.

(5)

1 INTRODUCTION

In this chapter the research problem of this study is introduced by addressing relevant background information on the research topic. After addressing the research problem, the problem statement and research questions will be discussed. Finally, an explanation of the contributing value of this study is explained, followed by a short outline of the following chapters.

1.1 Background research problem. In one of their studies, Keller & Lehman (2006)

claimed that already much academic research has been conducted with relation to the overarching research topic of this study; branding. But at the same time they also claim that there are many related research sub-topics of branding, that have not been addressed in

literature yet and therefore many research needs to be done with respect to branding according Keller & Lehman (2006). The high amount of attention that was already drawn upon branding in academic research, and the request of Keller & Lehman (2006) to conduct even more research on branding, seems logical since brands play an important role in the daily to daily life of almost each consumer and company. Many of the conducted studies on branding tried to set-up a set of rules on successful branding (E.g. Buchanan et al., 1999; Kay, 2005; Keller, 2008; Yoo, Donthu & Lee, 2000). Although there is no full agreement on the guidelines of branding, one of the central themes in these “Branding rules” is the consistency of brands. This “Consistency guideline” is applicable on several elements of branding. One of these elements is the brand’s positioning over time, which should be consistent according Buchanan et al., (1999). In light of this study, consistent branding refers to consistency of a brand in its positioning towards the chosen target group, which is shown in advertisements.

Several studies addressed the fact that brand consistency leads to a good brand image and higher brand equity (Buchanan et al., 1999; Keller, 1993). Especially the relationship between brand consistency and brand equity is important in light of this current study. This

(7)

doubts upon the influence of consistent branding (Kay, 2005; Brown et al., 2003; Charters, 2009). Regardless the doubts that exist in academic literature, this study examines the

relationship between consistent brand positioning, with reference to the target group shown in ads, brand attitude and attitudinal loyalty.

In examining the relationship between consistent brand positioning, with reference to the target group shown in ads, brand attitude and attitudinal loyalty, the question arises whether this relationship is static or influenced by third variables. It seems that there are some discrepancies in academic literature upon the influence of third variables on the construct of brand loyalty, of which attitudinal loyalty is a part. While, Grisaffe & Nguyen (2010) based on Oliver (1999) stated in their study that “situational factors” have no influence on brand loyalty when an ultimate loyalty exists. Buchanan et al., (1999) concluded that situational factors do influence the relationship between brand consistency and brand equity, the latter of which brand loyalty is considered to be a part of. Likewise, Kay (2005) found in his study that brand loyalty is influenced trough situational- and contextual factors. Also in a study of Carman (1970) it was shown that personality characteristics, which could also be seen as situational influences, had an effect on the shopping process and store loyalty of a customer. These latter lines of reasoning of Buchanan et al., (1999), Kay (2005) and Carman (1970) that brand equity/loyalty is influenced trough third variables forms the basis of this current study. According the CBBE model (Keller, 2008) brand attitudes form one of the layers before brand loyalty, attitudinal loyalty. Since the constructs, brand attitude and attitudinal loyalty relate to each other, it is reasonable to assume that brand attitude is also influenced by third variables, likewise attitudinal loyalty is.

Several studies concluded that brands both contain a “functional”- as well as a “symbolic” aspect, likewise consumers have “utilitarian” purchase motives targeted at the “functional” aspect of the brand and somewhat more “hedonic” purchase motives, targeted at the

(8)

customers (Kressman et al., 2006; Ericksen & Sirgy, 1989; Malhotra, 1981; Malhotra, 1988; Park, Jaworski, & Maclnnis 1986; Sirgy 1982; Sirgy & Samli, 1985). Since both functional congruity and self-consistency can be considered as situational/context variables, both factors could also influence the relationship between inconsistent brand positioning, brand attitude and attitudinal loyalty. As explained, inconsistent brand positioning in this study is defined, with respect to different target groups, which are shown in brand ads. This inconsistency in brand positioning could influence the symbolic aspect of a brand, since by showing

inconsistent target groups, consumer might get confused about the brand’s personality or the brand’s users image. Therefore, it is expected that for consumers with a high level of self-consistency, the negative relationship between inconsistent brand positioning, brand attitude and attitudinal loyalty will be strengthened since the inconsistency might interfere with their high level of self-consistency. Nevertheless, it is expected that the inconsistency in

positioning could not influence the functional aspect of a brand, since no product

functionalities are changed. So, with reference to the functional aspect, it is expected that for consumers with a high level of functional congruity, the negative relationship between inconsistent brand positioning, brand attitude and attitudinal loyalty is weakened. Since the high functional congruity level might weight up for the negative effect of the inconsistent brand positioning with reference to the symbolic aspect.

1.2 Problem statement. All the addressed background information on the research

topic of this study led to the following problem statement. The main research question of this study is: “How does inconsistent brand positioning, with reference to the target group shown

in brand ads, influence brand attitude and attitudinal loyalty and how does the level of functional congruity and self-consistency influence this relationship?” From this main

research question, some sub-research questions are derived. Namely, “What is the influence of

inconsistent brand positioning, with reference to the target group of the brand shown in ads on brand attitude and attitudinal loyalty?”, “How does functional congruity influence the relationship between “brand consistency”, brand attitude and attitudinal loyalty?” And, “How does self-consistency influence the relationship between “brand consistency”, brand attitude and attitudinal loyalty?”

1.3 Contributing value. This study contributes to existing literature on branding for

(9)

2005). Charters (2009) attempt with his study to solve this discrepancy in literature by

questioning the fact whether brands need to be consistent. A main finding of his study is “It is not consistent content, but a consistency of quality and the ability to deliver up to and beyond expectations that is crucial for the brand.” (Charters, 2009: 288). Although, the study of Charters (2009) delivered very valuable insights on the current discrepancy in literature on “consistency in branding”, the study was mainly focused on consistency of the actual product and product attributes. This current study will deliver insights on another aspect of brand consistency as Charters (2009), namely inconsistent brand positioning, with reference to the target group of the brand which is shown in ads. As far as current knowledge exists, the effects of inconsistent brand positioning, brand attitude and attitudinal loyalty have not been studied yet. Except for the indirect way, trough examining the relationship between brand consistency and brand equity, of which attitudinal loyalty is considered to be a part of (Buchanan et al., 1999). Nevertheless, this study is first in examining the direct relationship between inconsistent brand positioning, brand attitude and attitudinal loyalty. Secondly, this study is the first study which examines whether functional congruity and self-consistency have a moderating effect on the relationship between inconsistent brand positioning, brand attitude and attitudinal loyalty next to their direct effects which were determined by Kressman et al., (2006). This study does not only contribute to academic literature, but also has practical implications for managerial purposes. The outcomes of this study advises brand-managers, communication managers and advertisers on the possibilities in brand positioning, with reference to target groups shown in ads. The study also advises managers whether their decisions on brand positioning, with reference to the target group shown in ads, should be influenced by the level of functional congruity and self-consistency of their consumers.

1.4 Outline of chapters. This report starts in chapter two with an outline of the

theoretical framework. This framework brings together opinions, theories and concepts about the research topic, resulting in proposed relationships, hypotheses and a conceptual model. In the third chapter, the research design of the study is explained by means of describing the research method, data collection and analysis methods. This design is followed in chapter four by addressing the research results by means of analyses and a discussion of the provided data in chapter five. Finally, after discussing the results of the study, limitations and

(10)

2 THEORETICAL FRAMEWORK

In this chapter the stated research questions from the introduction chapter will be addressed by means of literature. This literature review brings together opinions, theories and concepts about the research topic. The review finally results in several hypotheses and a conceptual model, in which the founded and/or proposed relationships are graphically displayed.

2.1 Brand consistency. Brands, described by Keller (2008: 2) as “Something that has

actually created a certain amount of awareness, reputation, prominence, and so on in the marketplace” are nowadays a common part of our society. Just determining the amount of brands which customers are confronted everyday shows the impact brands have. Keller & Lehman (2006) even state that brands are one of the most valuable intangible assets a firm has. The important role of brands in our society also explains the fact why branding became a top management priority in the last decade and why branding became a common research area of many academic researchers (Keller & Lehman, 2006). This interest in branding of

academic researchers resulted in several articles and books about branding, especially about branding principles and branding guidelines (E.g. Buchanan et al., 1999; Kay, 2005; Keller, 2008; Yoo, et al., 2000).

(11)

Buchanan et al., (1999) and Bridges et al., (2000) it seems that consistency in branding has several applications, like consistency among brand extensions, brand communications and brand positioning over time.

The application of the consistency guideline on brand positioning is of relevance for this study. The positioning of a brand is seen as, the place a brand has within the product category, with reference of how similar or how different a brand is perceived in comparison with other brands in the category (Sujan & Bettman, 1989). “The essence of brand positioning is that the brand has a sustainable competitive advantage or a unique selling proposition.” (Keller, 1993: 6). This competitive advantage could give consumers a compelling reason for buying a particular brand, according to Keller (1993). So, the positioning of a brand also has a strong relationship with the customer base of a brand, since when targeting at a certain competitive advantage also a certain customer group will be attracted. According to Sujan & Bettman, (1989) there are several possible positioning strategies, whereas Bhat & Reddy (1998: 32) claimed “Positioning a brand through a clear and consistent image-building campaign has been a cornerstone of brand marketing practice.”. Leading to the assumption that whatever positioning strategy is used, the strategy should be clear and consistent. When summarizing the literature on consistent brand positioning it leads to a confirmation that consistent image building is necessary for maintaining a consistent brand positioning, towards customers over time and that this can be achieved by consistency among the various elements of a marketing campaign (Bhat & Reddy, 1998; Keller, 1993; Buchanan et al., 1999). This conclusion was also confirmed by Buchanan et al., (1999), namely that with reference to the consistency in the brand’s positioning over time it is stated that also each part of the marketing mix should reinforce that positioning.

With reference to this study a consistent brand is defined as “A brand that is consistent in its

positioning, with reference to its chosen target group, which is represented in brand advertisements.”. Contradicting, an inconsistent brand is defined as “A Brand that is not consistent in its positioning, with reference to different chosen target groups, which are represented in brand advertisements.”.

Although many academic researchers claim the importance of consistency in brand

(12)

branding seems to have its limits. In his article Kay (2005: 747) actually concludes “Consistency is increasingly difficult to sustain.” and “The logic of building brand equity through consistency and reinforcement of brand meaning appears to have its limits.” (Kay, 2005: 752). A study of Brown, Kozinets, & Sherry (2003) addresses the trend of reviving old brands to a successful new launch. Reviving of brands means that certain elements of a brand are innovated and updated. The innovations and updates could be perceived in some cases as inconsistent branding. Nevertheless this trend called the “retro revolution” is described as a powerful management option by Brown et al., (2003) which is often really successful. In his book Keller (2008) expressed some criteria useful for choosing and designing brand elements to build brand equity. Brand equity is the construct which could be reinforced trough

consistent branding as described by e.g. Buchanan et al., (1999). The criteria of Keller (2008) for choosing and designing brand elements are memorability, meaningfulness, aesthetic appeal, transferability, adaptability and flexibility over time and legal and competitive protect ability and defensibility. All criteria seem straightforward except the adaptability and

flexibility over time criteria, since this could lead to the opposite effect of brand consistency over time.As also addressed in the introduction, Charters (2009) found in his study that it is not the consistent content that is crucial for the brand and moreover he found that the need for brand consistency is product specific. These findings lead to the consideration that certain brands might not always need to be consistent, which is also concluded by Charters (2009) himself. Namely by stating “Brand managers with a type of product for which consumers seek a kind of product variation, may actually want to build variation and inconsistency into their product to encourage such an interactive consumer response.” (Charters, 2009: 289).

Nevertheless, it should be stated that Charters (2009) study was only directed at one aspect of brand consistency, namely consistency of the actual product and product attributes.

When summarizing opinions of academic researchers on consistent branding, it seems that there is a discrepancy in literature on the importance of consistency in branding. While some academic researchers (e.g. Bhat & Reddy, 1998; Keller, 1993; Keller, 1998; Buchanan et al., 1999) state that consistency in branding is necessary for a strong brand, others (e.g. Kay, 2005; Brown et al., 2003; Charters, 2009) doubt upon this consistency guideline. With

(13)

necessary is consistent positioning of a brand, with reference to the target group of a brand which is represented in advertisements?

As addressed above, according several academic researchers, the advocates of consistent branding, consistency has an effect on building a brand image and brand equity (Buchanan et al., 1999; Keller, 1993). Although there are also some comments on the consistency guideline in branding as addressed above, e.g. Kay (2005); Brown et al., (2003), the two proposed concepts as results of brand consistency, brand image and brand equity, deserve some more explanation about their meaning and influence. Brand image is a research concept in which academic researchers have been interested since already a long time. Actually according to Padgett & Allen (1997: 50) “The concept of brand image has a rich history in the marketing literature back to the 1950’s.”. Because of the rich history of the concept there are also several definitions of the concept available. These definitions range from relatively broad to more specific definitions. According Padgett & Allen (1997: 50) “Generally, brand image functions to define the product for consumers and differentiate the firm's offering from competitive offerings”. Keller (1993: 3) describes brand image as “Brand image are perceptions about a brand as reflected by the brand associations held in consumer memory”. This set of

associations could be used by the consumer to differentiate a firm’s offering from competitive offerings in line with Pedgett and Allen (1997) their definition of brand image. As Keller (1993: 3) also explained “The favorability, strength, and uniqueness of brand associations are the dimensions distinguishing brand knowledge that play an important role in determining the differential response that makes up brand equity, especially in high involvement decision settings.”. With reference to this study brand image is defined as “Associations a consumer

has stored in memory of a certain brand, which differentiate that certain brand from its competitors”. (Pedgett and Allen, 1997; Keller, 1993).

(14)

brand equity can be defined as the “Incremental cash flows which accrue to branded products over unbranded products”. Others refer to brand equity as the incremental utility or value added to a product by its brand name (Yoo et al., 2000). And according Kamakura & Russel (1993) the definition of brand equity depends on the fact whether brand equity is viewed as the value of the brand to the firm or the value of the brand to the consumer. In light of this study brand equity is viewed as the value of the brand to the consumer. Although as

described, there are many definitions available of brand equity, academic researchers agree upon the fact that brand equity is a multidimensional construct. As described, Keller (1993) explains that brand equity consists of brand knowledge, which is made up of brand

image/brand associations. Nevertheless according Aaker (1991) and Aaker (1996)1st brand equity consists of even more elements, namely brand loyalty, brand awareness, perceived quality, brand associations, and other proprietary brand assets. According Yoo et al., (2000) researchers identified similar dimensions as Aaker (1991) and Aaker (1996)1st of brand equity, nevertheless the authors also investigated themselves what the elements are of brand equity. They conclude “Considering the various suggestions, we recognize perceived quality, brand loyalty, and brand awareness with strong brand associations as common dimensions of brand equity” (Yoo et al., 2000: 196). Brand equity is in this study defined based on the conclusions of Yoo et al., (2000), namely “Brand equity consists of perceived quality, brand

loyalty, and brand awareness with strong brand associations and serves as the utility of branded products over unbranded products.”. This study will especially focus on one of the

purposed elements of brand equity, namely brand loyalty.

2.2 Brand Loyalty. As described, brand loyalty is considered as one of the building

blocks of brand equity (Yoo et al., 2000). As a part of brand equity but also as a construct on its own, brand loyalty, received already much attention in academic research (E.g. Aaker, 1991; Chaudhuri & Holbrook, 2001; Dick & Basu 1994; Oliver, 1999; Yoo et al., 2000). A well-known definition of brand loyalty is the definition of Oliver (1999). Namely “Brand loyalty is a deeply held commitment to rebuy or re-patronize a preferred product/service consistently in the future, thereby causing repetitive same brand or same brand-set

(15)

& Basu (1994). Behavioral loyalty means that repeated purchases of the brand are made. (Chaudhuri & Holbrook, 2001). According to Dick & Basu (1994) the behavioral definition of loyalty alone is insufficient to explain how and why brand loyalty is developed, since it does not cover the reason behind repurchasing. Therefore attitudinal loyalty focusses more on commitment by unique value associations with the brand according Chaudhuri & Holbrook (2001). So when interpreting the studies of Dick & Basu (1994) and Chaudhuri & Holbrook (2001) it seems that behavioral loyalty follows from attitudinal loyalty.

According the study of Chaudhuri & Holbrook (2001) brand trust and brand affect

(16)

With reference to this study brand loyalty is defined as “A twofold construct consisting of

behavioral- and attitudinal loyalty. Behavioral loyalty is the tendency of repeat purchases of a brand by a consumer and attitudinal loyalty is the tendency of commitment towards a brand trough unique value associations by a consumer” (Oliver, 1999; Chaudhuri & Holbrook,

2001). Although brand loyalty is a twofold construct, this study will focus on one aspect of loyalty, namely attitudinal loyalty since as stated behavioral loyalty could be formed through attitudinal loyalty. Based on literature (Buchanan et al., 1999; Kay, 2005) it is assumed that brand loyalty, attitudinal loyalty, is not a static construct, but that the construct is influenced by third variables, like situational factors.

2.3 Brand attitude. According the CBBE model of Keller (2008) building a strong

brand consists of four steps. The steps of building a strong brand form a “branding ladder” from identity, to meaning, to responses and finally to relationships. These four steps can be transformed in six brand building blocks according Keller (2008). The building blocks of Keller (2008) form together a pyramid of building a strong brand. The top of the pyramid is the resonance stage, which exists partly of brand loyalty. As stated in the previous paragraph, brand loyalty is influenced by the consistency of a brand’s positioning. The layer under the resonance stage in the pyramid of Keller (2008) consists of brand judgments and brand

feelings. Especially one aspect, the judgments part, is important in light of this study. Namely, the quality aspect, which consists of brand attitudes. Brand attitudes can be seen as customer’s overall evaluations of a brand according Keller (2008). Also Berger and Mitchell (1989) stated that brand attitudes can be defined as evaluations of a product or brand.

(17)

also influences the brand attitudes of consumers. This expectation can be drawn since the consistency of brand positioning also influences brand loyalty as was stated in the previous paragraph. Nevertheless, Keller (2008) states that brands can take two routes to eventually arrive at the resonance stage, namely the “rational” and the “emotional” route. The rational route contains the path from quality judgments, attitudes, to resonance. While the emotional route contains the path from brand feelings to resonance. Since the path from brand attitudes to brand loyalty is only one possible path to brand loyalty, brand attitudes are treated in this study as a separate variable next to attitudinal loyalty instead of a mediating variable. The treatment of brand attitudes as a separate dependent variable in this study, is in line with a trend described by Berger and Mitchell (1989). Namely, “More frequently, attitudes have served as dependent variables in studies examining consumer’s responses to marketing stimuli.” (Berger & Mitchell, 1989: 269). With reference to this study brand attitudes are defined based on the definitions of Keller (2008) and Berger and Mitchell (1989). Namely, brand attitudes are in this study defined as “The overall evaluations of a brand”.

2.4 Functional congruity and self-consistency. In one of their studies, Park et al.,

(18)

& Prabhakar (1990), concluded that customers brand knowledge is divided in terms of functional- and symbolic brand associations.

It can be concluded, based on above addressed literature that brands contain both a functional- as a symbolic aspect (Bhat & Reddy, 1998; Zeithaml, 1988; Park et al., 1986; Sirgy, 1982; Mittal et al., 1990). The functional aspect can be defined as satisfaction of functional/product related needs and the symbolic aspect can be defined as satisfaction of prestige

needs/expression of users’ personality (Bhat & Reddy, 1998; Kressman et al., 2006; Park et al., 1986). By combining both the functional- and symbolic aspect in one brand, two types of consumer motives can be addressed at the same time by one brand. One of these motives is the utilitarian motive, described as the “Motivation to pursue products and services that serve their primary function best at the lowest cost” (Kressman et al., 2006: 957). The other

consumer motive which is addressed by a brand is described as a hedonic motive, namely the fact that many consumers purchase goods and services that act as a vehicle to express their identity (Aaker, 1996 2nd; Sirgy, 1986).

Since, brands can both serve utilitarian and hedonic motives of customers, it seems likely that consumer’s assessment of either the functional and/or symbolic aspect of a brand has

influence on the brand attitude and the level of attitudinal loyalty of customers. In line with this reasoning, the study of Sirgy, Johar, Samli & Claiborne (1991) stated that consumer behavior is formed through functional- and self-congruity. Functional congruity refers to the review of consumers on the functional aspect of a brand, so in line with the utilitarian motive of consumers. Functional congruity is defined as “Assessment of the brand by focusing on the

extent to which functional attributes of the brand matches the consumer's ideal or desired performance specifications. Performance specifications are consumer expectations regarding how the product should perform to accomplish the focal or central goal of the product.”

(19)

Self-congruity is defined as “the match between consumers' self-concept (actual self, ideal self, etc.) and the user image or “personality” of a given product, brand, store, etc.”

(Kressman et al., 2006: 955). Self-congruity plays a very important role in determining brand loyalty, both by direct and indirect effects according Kressman et al., (2006). The concept of self-congruity is according Kressman et al., (2006) guided by several self-concept motives, with reference to the actual self, ideal self, etc. Especially, one of the self-concept motives, the consumers need for consistency, motivates brand loyalty (E.g. Ericksen & Sirgy, 1989; Malhotra, 1981; Malhotra, 1988; Sirgy & Samli, 1985). The need for self-consistency is defined as “The motivation of people to behave in ways consistent with how they see

themselves, so consistent with their actual self.” (Kressman et al., 2006: 956). The need for

self-consistency could lead to the fact that people purchase/consume goods with a brand personality which is consistent with their actual self. Since it was shown that consumers their need for consistency motivates a positive brand attitude and brand loyalty (Kressman et al., 200), the question arises whether self-consistency also influences the relationship between inconsistent brand positioning, with reference to the target group which is shown in ads, brand attitudes and attitudinal loyalty.

2.5 Expected relationships and hypotheses. When summarizing the above addressed

opinions, concepts and theories of academic researchers, some assumptions can be drawn upon expected relationships for this study. One of the main relationships to be assessed in this study, is the relationship between inconsistent brand positioning, with reference to the target group shown in ads, and attitudinal loyalty.As addressed brands both contain a functional- and symbolic aspect (Bhat & Reddy, 1998). The inconsistent positioning of a brand, with reference to the different target groups which are shown in ads, could be perceived as being a change in the “symbolic” aspect of a brand. Changing the target group, which is shown in ads, does not change the functional aspects of the brand/product, but it could change the perceived expression of users personalities and/or brand personality, which is considered to be a

“symbolic” aspect of the brand. Although there is a discrepancy in the opinion of academic researchers on brand consistency, the relationship between inconsistent brand positioning and attitudinal loyalty is expected to be negative, since in general academic researchers have claimed that brand consistency leads to brand image and brand equity, brand equity of which loyalty is considered to be a part of. (Buchanan et al., 1999; Keller, 1993). These

(20)

Master thesis Ilse Zwaan 20 H1: Inconsistent brand positioning, by showing different target groups in ads, leads to a decrease in attitudinal loyalty, opposite to consistent brand positioning.

The second main relationship to be assessed in this study, is the relationship between

inconsistent brand positioning, with reference to the target group which is shown in ads, and brand attitudes. As explained, the form of brand inconsistency tested in this study, could be perceived as a change in the symbolic aspect of a brand. Kressman et al., (2006) claimed in their study that brand attitudes of consumers are influenced by the symbolic aspect of a brand. When following this line of reasoning of Kressman et al., (2006), the question arises whether the effect of inconsistencies in branding on attitudes is positive or negative. As stated, it is expected that the relationship between inconsistent brand positioning and attitudinal loyalty is negative, since in general academic researchers claimed that brand consistency leads to a better brand image and higher brand equity, brand equity of which attitudinal loyalty is considered to be a part of (Buchanan et al., 1999; Keller, 1993). As described, according Keller (2008), brand loyalty is the highest level of the CBBE pyramid. Brand attitudes form partly the stage under the resonance stage, of which loyalty is a part, for strong brands in the CBBE pyramid (Keller, 2008). Since it is expected that inconsistent brand positioning

negatively influences attitudinal loyalty it is likely to expect that the stage under brand loyalty in the CBBE model is also negatively influenced trough inconsistent brand positioning. These expectations lead to the second hypothesis of this study.

H2: Inconsistent brand positioning, by showing different target groups in ads, leads to a decrease in positive brand attitudes, opposite to consistent brand positioning.

(21)

not change in the inconsistent brand positioning with reference to this study, it is expected that for consumers with a high satisfaction level of the need for functional congruity the inconsistency is less interfering. So, high functional congruity is considered to weaken the negative relationship between inconsistent brand positioning and attitudinal loyalty. Likewise it is expected to weaken the negative relationship between inconsistent brand positioning and brand attitudes, since brand attitudes partly lead to attitudinal loyalty. These expectations lead to hypotheses three and four.

H3: The higher the level of functional congruity for a consumer the weaker the negative relationship between inconsistent brand positioning, by showing different target groups in ads and attitudinal loyalty.

H4: The higher the level of functional congruity for a consumer the weaker the negative relationship between inconsistent brand positioning, by showing different target groups in ads and brand attitudes.

As explained, it is assumed that the relationship between inconsistent brand positioning, with reference to the target group shown in ads, brand attitude and attitudinal loyalty is influenced by situational factors. Self-consistency could be seen as a situational factor, which influences brand attitudes and brand loyalty, according Kressman et al., (2006).Self-consistency is considered to strengthen the already negative relationship between inconsistent brand positioning, attitudinal loyalty and brand attitudes. This expectation is drawn upon a finding of Kressman et al., (2006). Namely, “People have beliefs about their own identities, values, lifestyles, preferences, and habits. Once their “self-theories” are established, they become highly motivated to protect them” (Kressman et al., 2006: 956/957). They conclude by

addressing the fact that major threats to self-theories account even for mental breakdowns and psychosis (E.g. Lecky, 1945).

(22)

This interference by the inconsistent brand positioning with the “self-theory”, which is not protectable for consumers, could have negative effects. (E.g. Kressman et al., 2006; and Lecky, 1945). This interference is probably experienced more negatively for consumers with a high level of self-consistency than for consumers with a low level of self-consistency. This expectation is drawn upon the fact that for consumers with a high level of self-consistency, a change in the “symbolic” aspect of the brand could damage the self-consistency to a larger extent than for consumers with a lower level of self-consistency. These expectations lead to hypotheses five and six.

H5: The higher the level of self-consistency with a brand the stronger the negative

relationship between inconsistent brand positioning, by showing different target groups in ads and attitudinal loyalty.

H6: The higher the level of self-consistency with a brand the stronger the negative

relationship between inconsistent brand positioning, by showing different target groups in ads and brand attitudes.

Based on the previous addressed theory and hypotheses certain expectations can be drawn upon the research topic of this study. The proposed relationships are graphically displayed in figure 1. Figure 1 Conceptual model Inconsistent brand positioning, by showing different target groups in ads

(23)

3 METHODOLOGY

In this chapter the methodology of this study will be discussed. The chapter starts with a description of the research design. The description is followed by an explanation of the

research stimuli. Furthermore, a description about the study participants, variables and scaling and an explanation of the experimental procedure will be addressed. This chapter concludes by describing the plan of analysis for this study.

3.1 Research design. The research method of this study, an experiment, can be

classified according Malhotra (2010) as a primary quantitative data research method focused on determining causal relationships. An experiment can be described as “The process of manipulating one or more independent variables and measuring their effect on one or more dependent variables, while controlling for extraneous variables” (Malhotra, 2010: 253). The experiment was chosen as the research method for this study, since its design suits very well the purpose of this study, namely examining whether inconsistent brand positioning, the independent variable, effects attitudinal loyalty and brand attitudes, the dependent variables. Next it was determined whether these relationships between inconsistent brand positioning, attitudinal loyalty and brand attitudes, were affected trough the level of functional congruity and self-consistency, the extraneous variables.

(24)

positioning on attitudinal loyalty and brand attitude of the other experimental group was a contingency or whether it was statistically proven. Based on the fact that only one

experimental group saw the consistent advertising set and that the other experimental group only saw the inconsistent advertising set, it can be concluded that this experiment had a between-groups design.So, the independent variable of this study is a nominal variable, which was manipulated by the researcher and divided the research participants in two experimental groups. This independent variable can therefore be described as a categorical variable, which is called a factor, according Malhotra (2010). As will be further addressed in following paragraphs, the extraneous variables of this study were instead all metric variables, called covariates according Malhotra (2010) and are measured on an interval scale. So, in total there were two experimental groups in this study since only the independent variable divided the research participants in two groups, one group which was exhibit to a consistent advertising set and one group which was exhibit to an inconsistent advertising set, with reference to the target group shown in ads.

As addressed above, an experiment is focused on determining causal relationships (Malhotra, 2010). Although this term of a causal relationship seems quite straightforward, the concept deserves some explanation. The scientific meaning of causality is that the occurrence of X increases the probability of the occurrence of Y (Malhotra, 2010). But before a causal

relationship can be determined three conditions must be met. Namely, concomitant variation, time order of occurrence of variables and elimination of other possible causal factors

(25)

interpreted. So, all three conditions for causality were met, leading to the conclusion that when a relationship would be found in this experiment this should be a causal relationship.

3.2 Stimuli. The research area for this study is the business to consumer product

industry, i.e. consumer goods. This industry was chosen since the business to consumer industry is more “visible” and examinable than the business to business industry, with respect to the positioning of businesses towards target groups in ads through mass media etc. The product industry was chosen instead of the service industry, since the third variables under consideration, especially functional congruity, were more suitable for consumer good characteristics than for services, in which the functionalities could be more abstract and intangible for the consumer. Within this industry of consumer goods a specific brand was chosen as product stimulus for this study. In examining the possibilities for a brand as a product stimulus, the requirements were that the brand contains indeed a functional- and symbolic aspect and that the brand has a clear target group in its usual advertisements. This latter recruitment of a clear target group was necessary, otherwise the target groups could not be manipulated throughout the experiment for inconsistent positioning, with reference to the target groups shown in brand advertisements. First, the choice was made to choose a car brand for this study. The focus on cars in this study was made based on the study of Kressman et al., (2006). In this study the authors claim “Automobiles are high in conspicuousness, and that cars are therefore likely to be evaluated using symbolic criteria.” (Kressman et al., 2006: 958). This former argument of Kressman et al., (2006) formed the basis of choosing a car brand for this study since one of the extraneous variable of this study, self-consistency, is highly related to the symbolic criteria. Cars are also used across a variety of situations, according Kressman et al., (2006) and also have different functional aspects, which suits well with measuring the second extraneous variable of this study, namely functional congruity.

(26)

referring to MINI in this study, the classical models, constructed under the product name MINI, are meant. Notable, is the fact that MINI is quite straightforward in its positioning towards its target group. Actually, in one of their commercials they show different types of people, which they label either as a MINI driver or as a non MINI driver (www.youtube.com). From the MINI advertisements and by means of additional information, the target group of MINI can be described as a quite young target group, consisting of mostly urban

professionals, around 20 to 35 years old. The target group is trend conscious, has a good sense of style and is quality oriented. Next, the target group members are characterized by their active lifestyle, lust for life, spontaneity and need for excitement and joy.

(www.bmwgroup.com 3th; www.worldpress.com; www.autotell.com)

As addressed in the previous paragraph, the experimental stimuli for this study were two different advertising series for MINI. The ad series were manipulated with reference to the target group which was shown in the ads. Both the consistent- and the inconsistent advertising set were constructed by the researcher of this study. In constructing both advertising sets a regular advertisement of MINI was used, in which only the shown target market was changed. In the consistent ads the regular target group members of MINI were shown and claimed they love their MINI. But in the inconsistent advertising set, MINI users, who love their MINI, were shown that don’t fit in the regular target market of the brand. Except for the target group there are no differences between the both ad series, like lay-out. This choice was made

consciously, since only then any observed difference in the research results, could be assigned tothe differences in the target groups. An example of both a consistent ad and an inconsistent ad is shown in figure 2. Moreover, both whole ad series are shown in appendix A.

Figure 2

(27)

To make sure that the two experimental advertising series, were actually perceived being different in their level of consistent brand positioning, a pre-test was conducted with 15 respondents (Malhotra, 2010). Only when the respondents would actually observe differences in the level of consistent brand positioning, with reference to the shown target groups in ads, the tested ad series would be used in the experiment. In the pre-test the participants was shown a classical ad of MINI in which the regular target group of MINI was shown. After showing this classical advertisement the participants were handed some general information about MINI in which also the general positioning, with reference to the target group was described. After reading this information about MINI the respondents were told that MINI wanted to set up a new advertising campaign and that they doubt between two different ad series. The respondents were told they were about to see these two advertising possibilities for MINI and that their opinion on both advertising series would be tested. First the respondents was shown the consistent advertising set. After showing this set, the respondents were asked to answer some questions. In these questions a covered question was proposed about the consistency of the ads, with reference to the shown target group. Next the inconsistent advertising set was shown and the participants were asked to answer the same questions again. Finally the respondents were asked to compare both ad series, in which also a covered question was proposed about the level of consistency.

The pre-test for identifying whether people perceived inconsistent brand positioning, with reference to the target group shown in ads, was conducted online. The respondents of the pre-test were 15 random participants acquired through social media. In total there were nine male respondents and six female respondents, with an average age of 24 years old. Of the

(28)

These conclusions about the level of consistent brand positioning, with reference to the target groups which were shown in the ads, were confirmed in a comparison question which was proposed to the respondents. The respondents were namely asked to argument which campaign, either the first consistent, or the second inconsistent, would be best suited for MINI. Only one respondent stated that the second campaign would be best suited and one respondent did not answer this question. All other 13 respondents chose the first, consistent ad campaign, to be best suited for MINI. Arguments for this choice of the first campaign were mostly related to the people who were shown in the ads. Some quotes of the respondents “The first, since it is young, refreshing and ambitious” and “The first since young people are exhibit and this suits better with the target group of MINI”. So the pre-test confirmed that the two types of stimuli were perceived being different in the level of consistent brand positioning, with reference to the target group which was shown in the ads.

3.3 Research participants. As described in the research design, this study contains two

experimental groups. Both experimental groups should have the same group size and should contain comparable group members to avoid any bias caused by the experimental groups. To decide upon the appropriate group size of the two experimental groups, a study of Fennis, Janssen and Vohs (2009) and a study of Kardes, Fennis, Hirt, Tormala and Bullington (2007) with both a comparable research design were taken as a guideline. In their study Fennis et al., (2009) performed six experiments of which four experiments also contained two experimental groups, likewise this experiment. Likewise Kardes et al., (2007) performed three experiments of which two experiments also contained two experimental groups. The number of

participants of these experiments were respectively 37, 39, 60, 84, 147 and 155 leading to an average number of 87 participants. This average of 87 participants led to 43.5 respondents per experimental group. This number was completed to 40 participants per experimental group for this study. The participants of both groups should be people that fitted the regular target group of MINI. So, professionals in the age category of 20 to 35 years old. Nevertheless, also

students were included in the research participants since they represent the future potential target group of MINI buyers, by probably becoming young professionals in the near future.

3.4 Variables and scaling. As explained in chapter two, this study contains one

(29)

manipulated and controlled by the researcher of this study. Therefore no further measurement scale was applied to the independent variable of this study. Nevertheless this was not the case for the dependent- and extraneous variables of this study, which were all four linear variables measured on a particular scale as will be explained in this paragraph.

One of the dependent variables of this study is attitudinal loyalty, a part of brand loyalty. As described in chapter two, brand loyalty was handled in this study as a two-fold construct, namely attitudinal- and behavioral loyalty. Kressman et al., (2006) also state in their study that literature reports on two types of loyalty measurements, indeed attitudinal- and behavioral oriented measurements. Attitudinal loyalty can be measured according Kressman et al.,

(2006) by focusing on planned commitment of the consumer. This type of measurement is in line with the definition of attitudinal loyalty of this study, namely “Attitudinal loyalty is the tendency of commitment towards a brand trough unique value associations by a consumer.”. Both, Bloemer & Kasper (1995) and Chaudhuri & Holbrook (2001) focused in their studies on the measurement of planned commitment of consumers. The study of Chaudhuri & Holbrook (2001) determined planned commitment of a customer by means of two direct statements on commitment for a pre-determined brand by the researchers. Bloemer & Kasper (1995) used a more extended and indirect measurement on planned commitment, by means of six statements, for which the respondents themselves were free to choose a brand. In this study, the brand under consideration, MINI, was pre-determined by the researcher and therefore not all the measurement statements of Bloemer & Kasper (1995) were applicable. Since the scale of planned commitment of Chaudhuri & Holbrook (2001) is really compact, the scale of Bloemer & Kasper (1995) was not totally applicable, and since both scales

(Chaudhuri & Holbrook, 2001; Bloemer & Kasper, 1995) had some overlap, a combination of both scales was suitable for the purpose of this study.

For measuring the level of attitudinal loyalty the participants were asked to answer on which level of the seven point Likert scale1 they can identify themselves with three statements upon planned commitment. The seven point Likert scale was chosen since its outcomes are good to administer and since participants are likely to understand how to use the scale (Malhotra, 2010). The statements are placed in Appendix B.

1

(30)

As described, the second dependent variable of this study is brand attitudes. The overall evaluations of a brand were tested in this study according the study of Berger & Mitchell (1989). “Attitudes were measured using two seven point semantic differential scales,

anchored by good/bad and like extremely/dislike extremely.” according Berger and Mitchell (1989: 273). These two semantic differential scales were considered being appropriate for measuring brand attitudes, since Berger & Mitchell (1989) showed in their study that the two scales had a high internal validity. The statements used to measure the brand attitudes of respondents towards MINI are placed in appendix B.

(31)

Self-consistency is defined as “The motivation of people to behave in ways consistent with how they see themselves, so consistent with their actual self.” (Kressman et al., 2006: 956). From this definition it becomes clear that to measure self-consistency both a measurement of the self-image of a customer and a measurement of the brand personality is needed. For measuring both self-image and brand personality, Aaker’s (1997) brand personality scale was applied. Aaker’s (1997) scale consists of 5 brand personality dimensions, containing in total 15 elements. See Appendix C for an overview. Participants were asked to indicate on a seven point Likert scale to what extent the personality characteristics (Aaker, 1997) are applicable on their self. After applying these elements on them self, participants were asked to apply these elements on MINI, when imagining MINI as a person. All statements regarding the self-image and brand personality are exhibit in Appendix C. The level of self-consistency of the participants was also determined by the theory of Kressman et al., (2006). Namely,

“Computing absolute difference scores between each brand personality rating and its corresponding self-image ratings.” (Kressman et al., 2006: 959). These scores were finally averaged across all personality attributes for the participants. Furthermore the scores were revised so a high score represents a high level of self-consistency. (Kressman et al., 2006; Sirgy, 1982).

3.5 Experimental procedure. The experiment was conducted online. Participants

received personal invitations for the experiment either by e-mail or trough social media. Once the participant decided to cooperate with the experiment, the respondent could click on a link, from which he/she was directed to the experiment. All participants were randomly assigned to one of the experimental groups, so either to the group with inconsistent brand positioning or to the group with consistent brand positioning. It was chosen to assign the participants

randomly to the experimental groups, since any observed differences between the participants in the experimental groups than would be due to the experimental manipulation (Fennis and Stoebe, 2010). Also noticeable is the fact that participants could not decide themselves which experimental group they attended, since the participants didn’t know that there were two experimental groups.

(32)

measurements, see appendix C.1 and C.4. After answering these questions the respondents was shown a regular commercial of MINI, in which the regular target group of MINI was shown. Following they were asked to start reading a general description about MINI. This description contained some background information about the company, information about the type of product under consideration and of course about the regular target group, with reference to the positioning of MINI. It was chosen to not only provide information about the target group and positioning of MINI in the introduction, but also some general information. This choice was made since it was necessary that the participants did not know the research goal of this experiment, to avoid a research bias. Overall the introduction story was provided to make sure that all respondents had at least the same level of background information on MINI, so lack of background information could not cause any bias.

After reading the introduction story about MINI, the respondents were asked to answer questions about the level of functional congruity and self-consistency towards MINI. These questions are exhibit in appendix C.3 and C.5. Next, the respondents were asked to pay attention to some ads of MINI. At this point the actual experiment started, since the two groups were now being presented with the two different types of ad series, either consistent or inconsistent. After the respondents were exhibit to the advertising campaigns, they were asked to solve a small puzzle. This puzzle had no direct added value for this study, but acted as a kind of distraction task to change the mind-set of the respondents before testing their loyalty towards MINI. This distraction task was provided since in a regular market situation

consumers are not very often directly confronted with loyalty issues, as purchasing a good, after perceiving an advertisement. So the distraction task for the respondents was added to simulate a market situation as realistic as possible. After solving the small puzzle, the respondents were asked to answer some questions regarding their loyalty towards MINI. It should be noticed that although the distraction task was added in the experiment, the respondents were relatively quickly tested on their attitudinal loyalty and attitudes towards MINI after perceiving the advertisements. Normally, especially the inconsistency in

(33)

After answering the questions on attitudinal loyalty and brand attitudes the respondents were asked to go to the last page of the experiment. On this last page the respondents were asked to indicate the consistency level of the advertisements they saw throughout the experiment. This last question was proposed as a manipulation check. After this manipulation question the respondents were asked if they had any comments and/or remarks left. If so, there was space in the experiment to post these comments and/or remarks. After this last question the

respondents received a thank you message for participation and finished the experiment. A short summary of the experimental design is shown in table 1. Furthermore the whole experimental design is exhibit in appendix D.

Table 1

Summary experimental design

Part Experiment Purpose Questions

Welcome Introduction to the study No questions Personal information Testing whether the respondent actually

belongs to the target population and measurement self-participant and ideal functional aspects, with respect to self-consistency and functional congruity.

Questions 1, 2, 3, 4, 5, 6.

Introduction MINI (Movie and text)

Introducing the participants to MINI, of which the introduction to the target group is a part of.

No questions

MINI personality and functions

Measuring the self-scores of MINI and testing the functional aspects, with respect to self-consistency and functional congruity.

Questions 7, 8, 9

Advertisements Manipulation of the dependent variable by showing different stimuli.

No questions Puzzle Distraction task Question 10 Opinion MINI Measurement attitudinal loyalty and brand

attitude

Questions 11, 12 Opinion consistency Manipulation check Question 13 Additional information Space for participants to share comments or

additional information

No Questions

3.6 Analysis methods. The results of the experiment of this study should be

constructed into four variables per participant namely; the level of attitudinal loyalty, brand attitude, functional congruity and self-consistency. All four variables were

(34)

variable per construct. As stated, functional congruity and self-consistency were determined as one variable per participant by calculating the absolute differences between the constructs, obtaining a mean score and by reversing the scores. In order to determine the level of

attitudinal loyalty and brand attitude for the participants, the internal consistency on the statements of both constructs was measured by means of applying Cronbach alpha. Blumber, Cooper and Schindler (2005: 485) stated in their study “Cronbach alpha has the most utility for multi-item scales at the interval level of measurement”. Both attitudinal loyalty and brand attitude were measured by several items on a seven-point Likert scale, and therefore it was chosen to use Cronbach alpha for measuring the internal consistency of these constructs.

(35)

4 RESULTS

Within this chapter the results of this study are discussed. The chapter starts with a description of the research sample. This description addresses some general sample information and discusses the representativeness of the sample. Furthermore, the validity and reliability of the study is discussed. Following, the statistical methods for constructing the research constructs are addressed. The chapter concludes by describing the research results.

4.1 Representativeness sample. In total 80 participants, 40 per experimental group

cooperated in the experiment. In total there were 45 male participants and 34 female

participants (1 missing). As stated, the participants should either fit the current target group of MINI, so professionals between 20 to 35 years old, or they should fit the potential target group of MINI. Furthermore, participants of both experimental groups should be comparable to each other. All participants of both experimental groups did fit the age condition since all participants were between the age of 20 to 35 years old. The average age of the participants is 24.89 years old, with a standard deviation of 3.939. A graphical representation of the age distribution of the participants is exhibit in appendix E. As shown in appendix E, the age distribution of this study is non-normal. For the experimental group of the consistent advertising set the average age was 24.82 years old. For the group with the inconsistent advertising set the average age was 24.95. An independent sample T-test was conducted to see whether the age distribution was significantly different for both groups. The mean difference of 0.13 points is not a significant difference, as the independent sample T-test showed (Sig. .885 > 0.05). So, both groups were age-wise comparable.

(36)

groups are exhibit in table 2. Furthermore 48.8% of the participants is single, 18.8% has a relationship but currently does not live together. In total 26.3% of the participants lives together with his/her partner or is married and the remaining 6.3% of all respondents lives together and has children. Moreover, only one participant has owned a MINI in the past, one participant owns a MINI at this moment and all others have never owned a MINI. Since, the participants fit the requirements of the current or future target group of MINI, it can be concluded that the experimental sample is representative for MINI’s target group.

Table 2

Comparison tests experimental groups

Variable Test Mean scores/

percentages Consistent Mean scores/ percentages Inconsistent Result Age distribution Independent Sample T-test

24.82 Years old 24.95 Years old No sig. difference (Sig. .885 > .05) So, comparable groups. Daily profession Pearson Chi-Square Student: 57.5% Job seeker: 7.5% Professional: 35% Student: 50% Job seeker: 2.5% Professional: 47.5% No sig. difference (Sig. .374 > .05) So, comparable groups. Perceived consistency Independent Sample T-test 5.10 points on consistency level on a 7 point Likert scale.

3.28 points on

consistency level on a 7 point Likert scale.

Sig. difference. (Sig. .000 < .05) So no comparable groups.

4.2 Validity and reliability study.

Overall, validity of a study addresses the fact

whether a study actually measures the things the study is supposed to measure. According to Malhotra (2010) validity of an experiment contains two aspects, internal validity and external validity. Where internal validity means “Whether the observed effects on the test units could have been caused by variables other than the treatment.” (Malhotra, 2010: 254). External validity means “Whether the cause-and-effect relationships found in the experiment can be generalized.” (Malhotra, 2010: 255). First of all, to test the validity of this study a

manipulation check was conducted. This check measured whether the manipulated stimuli, the advertising series, indeed had the effect they were supposed to have, namely perceived consistency and perceived inconsistency. The manipulation check was performed at the end of the study, by asking the participants to indicate how consistent they thought the

advertisements they saw were on a semantic difference scale. As was expected, the

(37)

respectively 5.10 and 3.28 on a seven point Likert scale. The mean difference of 1.82 points is a significant difference, as an independent sample T-test showed, see table 2 (Sig. .000 < 0.05). Furthermore, it can be concluded that the internal validity for this study is high. As similarly addressed in the previous chapter, with respect to the causality of the examined relationships, only the target group in the stimuli was changed for both experimental groups. Consequently, effects on the dependent variables of the study could only be caused by this particular change in the stimuli. Moreover, also other factors that could cause an effect were controlled for, since the extraneous variables in the experiment were also measured and interpreted.

With respect to the external validity, it is noticeable that the examined relationships of this study were tested with only one type of product and one certain target group, respectively a car and the target group of MINI. The conclusions of this experiment can therefore only be transferred to comparable consumer goods and comparable target groups as MINI. This issue will be addressed more-in-depth in the generalizability section of this study. To make the study as reliable as possible, the experiment was set-up to reflect the market situation as realistic as possible. The simulation of the market was performed by e.g. the distraction task for the participants. By stimulating the market as realistic as possible, future experiments concerning this research topic should have comparable outcomes as this experiment. Therefore this experiment can be considered as reliable.

4.3 Determination constructs. As stated in the previous paragraph, in order to apply

the statistical methods necessary for this study, four multiple constructs needed to be reduced to single variables. For the level of self-consistency, the scores of the personality traits of MINI were subtracted from the scores of the participants’ personality traits. For the level of functional congruity the ideal scores for car attributes were subtracted from the scores of MINI per participant. Theses difference scores were handled as absolute values. Per participant all difference scores, for both self-consistency and functional congruity, were averaged. After averaging, the scores with a low value represented a high level of

Influence of inconsistent brand positioning on brand attitude & attitudinal loyalty