Brand Associations:
How the Size of Associative Networks influences its Content
University of Amsterdam
Supervisor: Drs. Jorge Labadie MBM Second Reader: Drs. Roger Pruppers Final – 15. August 2014
Student: Anna Seidel Student ID: 10599568
Master of Science Business Studies Track: Marketing
TABLE OF CONTENT ! ABSTRACT ... 1! 1.! INTRODUCTION ... 1! 1.1.! PROBLEM STATEMENT!...!2! 1.2.! CONTRIBUTIONS!...!5! 1.2.1.! THEORETICAL CONTRIBUTION!...!5! 1.2.2.! MANAGERIAL IMPLICATIONS!...!5!
1.3.! STRUCTURE OF THE THESIS!...!6!
2.! CONTENT OF ASSOCIATIVE NETWORKS ... 7!
2.1.! AAKER’S CATEGORIZATION OF BRAND ASSOCIATION!...!8!
2.2.! KELLER’S CATEGORIZATION OF BRAND ASSOCIATION!...!10!
2.3.! CATEGORIZATION OF BRAND ASSOCIATION WITHIN THE BRAND RESONANCE PYRAMID!...!12!
3.! STRUCTURE OF ASSOCIATIVE NETWORKS ... 15!
3.1.! HUMAN ASSOCIATIVE MEMORY THEORY AND SPREADING ACTIVATION PROCESS!...!15!
3.2.! KNOWLEDGE ACTIVATION AND ACCESSIBILITY!...!16!
3.3.! KNOWLEDGE ACTIVATION ORDER!...!17!
3.3.1.! Category Associations!...!18! 3.3.2.! Performance Associations!...!18! 3.3.3.! Imagery Associations!...!19! 3.3.4.! Judgements!...!20! 3.3.5.! Feelings!...!20! 3.3.6.! Resonance!...!21!
4.!FACTORS INFLUENCING THE SIZE AND THE CONTENT OF ASSOCIATIVE NETWORKS ... 22!
4.1.! CONSUMER EXPERTISE!...!22!
4.2.! BRAND USAGE AND FAMILIARITY!...!23!
4.3.! PURCHASE INTENSION, CATEGORY INVOLVEMENT AND ATTITUDE!...!24!
4.4.! SIZE OF ASSOCIATIVE NETWORKS!...!25!
5.! HYPOTHESES DEVELOPMENT ... 26!
6.! DATA AND METHOD ... 31!
6.1.! STIMULI DEVELOPMENT!...!31!
6.3.! CATEGORIZATION PROCEDURE!...!36! 6.4.! DESCRIPTION OF VARIABLES!...!38! 6.4.1.! Dependent Variables!...!38! 6.4.2.! Independent Variables!...!38! 6.5.! CONTROL VARIABLES!...!39! 6.6.! MODEL SPECIFICATION!...!40! 7.! RESULTS ... 43! 7.1.! MANIPULATION CHECK!...!43! 7.2.! SCALE RELIABILITY!...!43! 7.3.! DESCRIPTIVE STATISTICS!...!45! 7.4.! CORRELATION ANALYSIS!...!47! 7.5.! REGRESSION ANALYSIS!...!50!
7.5.1.! REGRESSION ON THE NUMBER OF COVERED CATEGORIES!...!50!
7.5.2.! REGRESSION ON THE PROPORTION OF ASSOCIATIONS RELATED TO THE “SALIENCE” CATEGORY!...!52!
7.6.! ADDITIONAL ANALYSIS!...!55!
7.7.! QUALITATIVE ANALYSIS!...!57!
8.! DISCUSSION ... 62!
8.1.! ADDITIONAL FINDINGS OF INTEREST!...!64!
8.2.! QUALITATIVE FINDINGS!...!66!
8.3.! MANAGERIAL IMPLICATIONS!...!67!
8.4.! LIMITATIONS AND FUTURE RESEARCH!...!67!
9.! CONCLUSION ... 69!
REFERENCES ... 71!
LIST OF TABLES AND FIGURES
TABLE 1: TYPES OF BRAND ASSOCIATIONS ... 7
TABLE 2: PRE-TEST 1(BRAND RECALL) ... 32
TABLE 3: PRE-TEST 2(BRAND CONCEPT) ... 33
TABLE 4: HIERARCHICAL MULTIPLE REGRESSION MODELS ... 42
TABLE 5: DESCRIPTIVE STATISTICS AND CORRELATION MATRIX ... 48
TABLE 6: REGRESSION ON THE NUMBER OF COVERED CATEGORIES ... 52
TABLE 7: REGRESSION ON THE PROPORTION OF "SALIENCE"ASSOCIATIONS ... 54
TABLE 8: ADDITIONAL CORRELATION MATRIX ... 61
TABLE 9: REGRESSION ON THE TOTAL NUMBER OF PERFORMANCE ASSOCIATIONS ... 61
FIGURE 1: DIMENSIONS OF BRAND KNOWLEDGE ... 11
FIGURE 2: BRAND RESONANCE PYRAMID ... 13
FIGURE 3: WORD CLOUD OF NIKE (SMALL ASSOCIATIVE NETWORK) ... 57
FIGURE 4: WORD CLOUD OF ASICS(SMALL ASSOCIATIVE NETWORK) ... 58
Abstract
It is generally accepted in the branding literature that consumers store brand knowledge in form of associative networks in their memory. Past research has shown that the size of these associative networks is an important association characteristic and has a positive influence on (customer-based) brand equity (Krishnan 1996; Chen 2001). Brand usage and consumers’ expertise on a product category are found to increase the total number of brand associations while simultaneously having an influence on the content of association. This thesis aims at examining the relationship between the size of associative networks and its content. Using a free association procedure, brand associations were elicited and categorized into different types of associations. After controlling for brand usage and consumer’s expertise, the influence of the size of associative networks (measured by the total number of associations) on the content (measured by the proportion of different association categories) was investigated. For this purpose one functional and one symbolic brand belonging to the same product category were used as stimuli. It was argued that small associative networks differ from large associative networks in terms of their content. The statistically significant results show, that the associative networks evolves to be more multi-dimensional with an increase in the total number of associations, however, little evidence could be found that the size of the associative networks has a significant influence on the elicited content. Nevertheless, interesting additional results were revealed. These findings contribute to the understanding of associative networks and multiple factors which influence them.
1. Introduction
Imagine a scenario in which a consumer has the intention to buy a new car in the near future. Different automotive brands such as Volvo, Toyota, Volkswagen, Audi, BMW, Porsche, or Bentley might pop up in his mind. “Porsche and Bentley are very expensive”, the consumer might think. So he narrows down his consideration set to Volvo, Toyota, Volkswagen, Audi and BMW. He considers multiple product attributes and their benefits to be able to compare different models, form a judgment and make a purchase decision. Additionally, thoughts like “Volkswagen, Audi and BMW are German cars, they are very reliable” or “I really liked the
advertising from Toyota I saw the other day” might go through his head. His decision could
also be influenced by positive or negative experiences he had with a brand or by feelings, which get evoked when he imagines himself driving a certain car.
This fictitious scenario describes a decision-making process in which the consumer retrieves brand knowledge stored in memory. This knowledge is stored in form of complex networks consisting of associations linked to a brand name and to one another. These networks are referred to as “associative networks”. With a large investment such as purchasing a new car, consumers are likely to think actively about what they know about a brand. A lot of different brand associations might pop up in their heads. These associations have an impact on cognitive considerations of product benefits and can elicit positive affect. This in turn influences consumers’ evaluations, purchase decisions and brand loyalty (Alba et al. 1991; Van Osselaer and Janiszewski, 2001).
However, there are also situations, in which consumers might not be able to process and recall all the information they store in their memory about a brand. When a consumer finds himself in the supermarket in front of a shelf of different washing detergents, he is less
likely to make the cognitive effort to take all the available information into consideration. This can be due to a lack of motivation because he doesn’t care that much about which brand to choose as long as it fulfils its purpose. Or he doesn’t have the ability to process the information because he is in a hurry, for example. In this case, only a few associations might pop up in his mind.
If only a few associations get recalled from memory, which types of associations will that be? They could be related to the product category and the need satisfaction such as: “Washing detergents remove stains from clothes”. The associations could also be related to different products, judgements about their effectiveness or they could be related to the feelings which get evoked by the smell of a washing detergent.
The question arises whether the same types of associations get activated when only a few associations are recalled from memory compared to when a lot of associations pop up in consumers’ minds. Or do the types of associations differ from each in these two scenarios?
What goes on in the minds of consumers is hidden and oftentimes unconscious. Even though multiple scholars tried to bring light into the darkness of consumers’ minds, a lot of questions still remain unanswered. It is important to gain a better understanding about which types of associations get recalled from memory and which factors influence the retrieval of brand knowledge since it has an influence on a wide range of consumer behaviour.
1.1. Problem statement !
Given the importance of understanding the network of associations in consumers’ minds, this provoked a large research stream about various characteristics of brand associations and associative networks.
Several authors took a closer look at the content of associative networks and classified brand associations into different categories (e.g. Aaker 1991, Keller 1993, Korchia 2000). Brandt et al. (2011) found that these networks may differ from one individual to another depending on the cultural background and/or the experience with the brand. Oakenfull and McCarthy (2010) researched how brand associations differ from each other depending on the level of brand usage and Czellar and Luna (2010) investigated the influence of expertise. Brand usage and consumers’ expertise about a product category are found to increase the total number of brand associations and hence the size of associative networks while simultaneously having an influence on the content of association.
According to Krishnan (1996), the size of associative networks is an important association characteristic. When consumers are asked what comes to mind, when they think about a certain brand, some consumers will have a large associative network and elicit multiple thoughts, while others have a small associative network and can only mention one or two associations. Krishnan (1996) proposed that, from an brand equity perspective, it is important to have a large number of associations. Chen (2001) supported this, after finding a significant positive effect of the total number of brand associations on customer-based brand equity. Brand equity describes a set of assets linked to a brand’s name and symbol, which creates value to the firm and its customers (Aaker 1991). These assets include brand loyalty, brand awareness, perceived quality and brand associations (Aaker 1991). Chen (2001) argued that compared to the other assets, brand associations are the core asset for building strong brand equity.
Meyers-Levy (1989) found that the size of the associative network has an influence on the memory of brand names, categories and message information. The author argued that with an increase in the number of associations, the memory structure becomes more complex. Generally, this makes it easier to recall brand information from memory since these
associations offer multiple pathways to the same brand node. When the size of associative networks has an influence on which information get retrieved from memory, it is reasonable to suggest that a causal relationship between the size of the associative network and the elicited content could exist. To this date, scholars have focused on investigating either the size of the associative network or took a closer look at the content. However, nobody has examined the relationship between these two characteristics. Therefore, the goal of this master thesis is to investigate this topic, address the gap in the literature and to examine the following question: “What influence does the size of the associative networks have on the
content of associative networks?
To be able to properly examine this topic, several underlying subtopics will to be discussed: 1. Content of associative networks:
Which different types of associations can be linked to a brand name and how can they be categorized?
2. Structure of Associative networks:
How are associations stored in consumer’s memory?
How do associations get activated and retrieved from memory? Which associations are easier to retrieve than others?
3. Influencing factors:
Which variables have an influence on a) the size of associative networks and b) on the content?
1.2. Contributions
1.2.1. Theoretical contribution
The findings of this thesis will add to the existing literature of associative network models by examining the relationship between two important association characteristics, the size and the content of associative networks. These two characteristics individually have been studied extensively, but even though previous research suggests that the size has an influence on memory structure and information processing (Meyers-Levy 1989), how the size of associative networks influences the content hasn’t been investigated before. This thesis lays a focus on this relationship. Apart from that, it draws on findings and theories from cognitive science and the existing information processing literature in order to make sense of which hidden processes are going on in the minds of consumers.
1.2.2. Managerial implications
Understanding the associations a brand evokes in the mind of consumers provides valuable insights to companies (Aaker, 1991). According to Keller (2009, p. 48) “[…] the power of a
brand lies in the mind of consumers and the meaning that the brand has achieved in the broadest sense”.
Marketers and brand managers are strongly interested in brand association networks because not only they identify vital brand associations but they also visualize how they are connected to the brand and to one another (John et al. 2006). Establishing key brand associations in the mind of consumers is essential for brand positioning in order to differentiate the brand and create a competitive advantage (Keller and Lehmann, 2006).
The insights of this thesis will be valuable information to brand managers and marketers as they provide a more detailed picture of the associations consumers hold in memory about their brand, which enables them to better understand what drives attitudes and how to influence purchase decisions. It reveals areas of improvement for brands by showing
whether brand managers should focus on increasing the numbers of associations by making consumers more knowledgeable about a brand or focussing on the attempt to strengthen a specific type of brand associations in the mind of consumers. This understanding is crucial as is has an impact on customer-based brand equity (Krishnan 1996, Chen 2001).
1.3. Structure of the thesis
The following section will elaborate on the existing literature about associative networks, its content and structure. This knowledge combined with findings from cognitive science form the foundation to develop several hypotheses. Two pre-tests will be conducted to identify one functional and one symbolic brand used as stimuli in the main study. In the main study, associations will be collected through a free association procedure and afterwards categorized into different association categories. The hypotheses are tested with different multiple regression models. In the end, the findings will be discussed and conclusions implications will be drawn.
2. Content of Associative Networks
“A brand association is anything linked in memory to a brand” (Aaker 1991, p. 109), including thoughts, feelings, perceptions, images, experiences and so on (Keller, 2009). Given that these associations can vary broadly and are not equally important in terms of directly or indirectly affecting buying behaviour, several authors made the attempt to classify associations in various ways, which is depicted below. A few of these alternative categorizations of brand associations will be discussed in greater detail.
Table 1: Types of Brand Associations
Literature Product associations Organizational associations
Functional attribute Non-functional attribute Corporate ability Corporate social responsibility Aaker (1991) Product attributes
Rational customer benefits
Product class
Intangibles
Psychological customer benefits Relative price Use/application User/customer Celebrity/person Life-style/personality Country/geographic area
Biel (1992) Functional product
attributes Soft or emotional product attributes Functional corporate attributes Soft or emotional corporate attributes Farquhar and
Herr (1993) Product category Product attribute Customer benefits Usage situation Customer benefits Keller and Aaker (1995) Innovativeness Environmentally conscious Community minded
Aaker (1996) Perceived quality
Innovation
Presence and success Local vs. global
Society/community orientation
Concern for customer Chen (1996) Perceived quality
Functional feature Symbolic Emotional Innovativeness Brown and Dacin (1997)
Corporate ability Corporate social responsibility
Korchia (2000) Product category Product-related attributes Functional benefits Brand personality Celebrities and events Users Usage situations Price Communication Distribution Experiential benefits
Symbolic benefits and attitudes
Company
Other organizations
Keller (2001) Product category Need satisfaction Primary characteristics and secondary features Product reliability Durability Efficiency Style and Design Price
Service User Profiles
Purchase and Usage situations Personality and values History, heritage and experiences Judgements Feelings
(based on Chen 2001)
!
2.1. Aaker’s Categorization of Brand Association !
Aaker (1991) identifies eleven types of associations that are linked to a brand, its name, symbol and slogan, which can be important associations in itself. These associations include product attributes, intangibles, customer benefits, the relative price of a product, its use or application, the typical user or customer, celebrity endorsers, the life-style or personality associated with a brand, its product class, the brand’s competitors and the geographic area or country that a brand comes from.
When a consumer is given a brand as a cue, it often elicits thoughts about the product
attribute or characteristics. These associations can directly be translated into reasons to buy a
brand, given that the product attribute is meaningful (Aaker 1991).
Intangible factors summarize sets of more objective attributes, such as perceived
quality, technological leadership or perceived value. According to Aaker (1991) it is easier for consumers to retrieve intangible information from memory, such as a cereal being “healthy”, instead of processing detailed product attributes like the list of ingredients (Aaker 1991). Intangible associations can also be used on a corporate-level by transferring
associations such as innovativeness or perceived quality to products under the brand’s umbrella (Aaker 1991).
Customer benefits can be divided into rational benefits and psychological benefits
(Aaker 1991). While rational benefits are closely related to a product attribute, psychological benefits are linked the feelings which are evoked during or after the brand experience. Consequently, rational benefits are part of a “rational” decision process, whereas psychological benefits involve an attitude-formation process (Aaker 1991). For example, the rational benefit of head&shoulder’s shampoo is getting rid of dandruff, but the psychological benefit is the consumer’s supposed enhanced confidence. Aaker (1991) states that a psychological benefit can be a powerful form of association and will be more effective when it is grounded in a rational benefit.
Furthermore, a brand can be strongly associated with its relative price, which is why Aaker (1991) considers this product attribute separately. When consumers think about IKEA for example, one of the first associations, which come to their mind is probably “cheap” furniture and Porsche will most likely be associated with “expensive” sports cars.
The use or application of a brand is another group of associations. When consumers are given Kellogg’s cereal as a stimulus they will most likely think about breakfast or the start of the day. In addition, brands can also be linked to its users in consumers’ mind. In the case of Harley Davidson, the brand will probably also evoke thoughts about Harley Davidson motorcycle riders. According to Aaker (1991) a strong association with the user group can be positive, however, it can also limit a brand’s ability to expand into new market segments.
A lot of companies use celebrity endorsers to promote their products because they leave a strong association in consumer’s memory (Aaker 1991). Nike used the basketball player Michael Jordan as an endorser in its advertising and Krishnan (1996) found Michael Jordan as one of the key associations in Nike’s associative network. However, a brand does
not necessarily have to be associated with a celebrity, as the example of the Marlboro man illustrates. The identity and characteristics that this fictional character conveys have become decisive associations for the brand (Aaker 1991).
In some cases, brands can be associated with personality and life-style traits (Aaker 1991). Brand personality attributes may be formed based on user and usage imagery and are defined as “a set of human characteristics associated with a brand” (Aaker, 1997). For instance, the personality characteristics associated with Coca-Cola are “cool”, “all-American” and “real” (Aaker, 1997). In multiple studies, brand personality has been shown to have an important role in creating brand knowledge (e.g. Kapferer 1992, Aaker 1997). Additionally, associations about a brand may also reflect the product-class in which a brand is positioned (Aaker 1991).
Sometimes when consumers think about a brand, thoughts about competitors will be evoked as well (Aaker 1991). This was the case in Krishnan’s study (1996) where Nike was associated with athletic shoes, which in turn was linked in consumers’ memory to Reebok, one of Nike’s competitors.
Lastly, Aaker (1991) argues that a brand may conjure an association with a country or
geographic area. Absolut Vodka is associated with Sweden (Aaker 1991) and Levi’s and
Marlboro are liked to America, just to name a few examples.
2.2. Keller’s Categorization of Brand Association
Keller (1993) classifies brand associations into nine dimensions, comprising three main categories with increasing scope: attributes, benefits, and attitudes.
Attributes are descriptive features, which characterize a product or a service and can
Figure 1: Dimensions of Brand Knowledge
(Keller, 1993)
Product-related attributes enable the performance of a product or a service, whereas non-product-related attributes are external aspects relating to its purchase or consumption such as the price, packaging, user group and usage situations (Keller 1993).
Benefits are defined as the personal value consumers attach to the product or service
attributes or in other words, what consumers expect the product or service can do for them (Keller 1993). Depending on their underlying motivations, benefits can be subdivided into functional, experiential and symbolic benefits (Park et al. 1986). Functional benefits and
experiential benefits both belong to product-related attributes, but the former arises from the
intrinsic advantages of the use of a product or service, while the latter corresponds to what a consumer feels during product or service consumption (Keller 1993). Symbolic benefits on the other hand relate to non-product related attributes and satisfy consumers’ needs for social recognition and self-expression (Solomon 1983).
Brand Knowledge Brand Image Brand Awareness Types of Brand Associations Favorability of Brand Associations Strength of Brand Associations Uniqueness of Brand Associations Brand Recall Brand Recognition Attitudes Benefits Attributes Functional Symbolic Experiental Product related Product related Price User/ Usage Image Brand personality Feelings & Experience
The last category consists of brand attitudes, which are defined as consumer’s overall evaluation of a brand (Wilkie 1986). Brand attitudes are crucial, because they often represent the basis for consumer behaviour such as brand choice (Keller 1993). Do note that Keller’s (1993) classification can be compared to Aaker’s (1991) first three association categories as Keller’s (1993) attitudes can be treated as equal to Aaker’s (1991) intangibles.
2.3. Categorization of Brand Association within the Brand Resonance Pyramid
In 2001, Keller introduces the Brand Resonance Pyramid, a model to build a brand and customer-based brand equity (CBBE) in four steps. The Brand Resonance Pyramid is another way to classify consumer’s associations with brands.
The first level of the pyramid is brand salience. It ensures that consumers associate a brand with a specific product class, product benefits, or customer needs. At the second level consumers should hold tangible or intangible brand associations in their minds. This level includes rational associations about the performance of a brand or a product (such as primary product characteristics, reliability, durability or the price) on the left side, as well as more emotional associations about the imagery (e.g. user profiles, purchase and usage situations, brand personality and values, heritage and experiences) on the right side. Associations related to brand performance describe how well the product or service satisfies consumers’ more functional needs. Brand imagery on the other hand, refers to the more extrinsic properties of a product or service with which consumers try to satisfy their psychological or social needs. These brand associations are more abstract and intangible and can result from direct experiences or indirectly through advertising or word-of-mouth. The third level of the pyramid, or the third step in building CBBE, reflects consumer responses to the brand, their overall judgments about quality, credibility and superiority on the left side and feelings that consumers associate with the brand on the right side. Brand judgments are personal opinions
about a brand. Consumers evaluate a brand by putting together the brand performance and brand imagery associations. Brand feelings on the other hand mirror consumer’s emotional responses and reactions evoked by a brand. These primary and secondary associations and brand responses serve as a foundation to create brand resonance on the fourth level of the pyramid. Associations related to brand resonance describe a deep psychological bond between the consumer and the brand (Keller, 2001).
Figure 2: Brand Resonance Pyramid
!
(Keller, 2001)
In addition to the horizontal division, brand associations can also be divided vertically into functional associations on the left side of the Brand Resonance Pyramid and symbolic associations on the right side. Park et al. (1986) propose that every brand is based on a brand concept or a brand specific-abstract meaning, which can be either symbolic or functional. The brand concept influences which types of associations a brand induces.
Resonance Loyalty Attachment Community Engagement
!!
!!
Judgments Quality Credibility Consideration Superiority ! Feelings Warmth Fun Excitement Securoty Social Approval Self-Respect PerformancePrimary Characteristics and Secundary Features
Produc Reliability Service Effectiveness, Efficiency, and Empathy
Style and Design Price
Imagery
User Profiles Purchase and Usage
Situations Personality and Values History, Heritage, and
Experiences Salience Category Identification Needs Satisfied Functional associations Smymbolic associations
Functional brands satisfy practical needs related to specific and practical consumption problems, whereas symbolic brands satisfy the need for self-image and social identification. Also their practical usage only has a subordinate role (Bhat and Reddy 1998). For instance, to owners of Harley-Davidson motorcycles their motorcycles mean more than just the pure mode of transportation. Rather, they associate an experience, an attitude and a life-style with using the brand. However, in comparison to the Harley-Davidson buyers, owners of Honda motorcycles most likely don’t see a symbolic value in their motorcycles which serve only for a functional reason: transportation. Even though, Harley-Davidson and Honda can be found in the same product class, they have a different brand concept. While Harley-Davidson is a symbolic brand, Honda is perceived as a functional brand (Bhat and Reddy 1998). The findings of the conducted study of Bhat and Reddy (1998) suggest, that brand functionality and symbolism are distinct concepts in consumers’ minds.
Therefore, it can be assumed that the content of the associative networks of functional brands differs from the content of brands with a symbolic brand concept. Functional brands are more likely to elicit thoughts and associations concerning the brand performance and rational, overall judgments of the brand. Symbolic brands on the other hand, are likely to have more associations that can be categorized under imagery as well as feelings attached to a brand. Looking at Keller’s Brand Resonance Pyramid, functional brands will follow the “rational route” on the left side of the pyramid, while symbolic brands follow the “emotional route” on the right side to achieve resonance.
3. Structure of Associative Networks
3.1. Human Associative Memory Theory and Spreading Activation Process
It is widely accepted that consumers store brand information in their memory in form of networks (e.g. Anderson and Bowler 1973, Collins and Loftus 1975, Krishnan 1996). The research stream on “associative network models” has evolved from psychology and cognitive science and can be traced back to Anderson and Bower’s (1973) Human Associative Memory (HAM) theory.
HAM theory assumes that semantic memory or knowledge consists of a set of nodes that are linked together to form a complex network (Collins and Loftus 1975, Anderson 1983). These nodes contain information about a brand and can be recalled from memory through a “spreading activation” process from one node to other liked nodes (Collins and Loftus 1975). Consider the example of McDonalds. The brand name can elicit a strong association with fast food, which is linked in consumer knowledge with convenience, drive through, or the perception of being unhealthy. In turn, the association of unhealthy food can stimulate the recall from memory of greasy and fatty fries or the cause of obesity (French and Smith 2013). Anderson (1983a, p. 86) refers to this process as a “chain of reaction, which provides the
energy that runs the cognitive machinery”.
Krishnan (1996) defines the links between two nodes as an “association” in the minds of consumers. Consumer brand associations are defined as those perceptions, preferences and choices that are linked to a brand in consumers’ memory (Aaker 1991). These associations can range from physical product attributes, usage situations or brand logos to perceptions of people, places, and occasions (Henderson et al. 1998; John, et al. 2006).
Initially, a brand may only be associated with its products, however over time and through marketing activities or product experiences a large set of associations can developed
in the heads of consumers that exist beyond the objective product offering (Keller and Lehmann 2006).
3.2. Knowledge activation and accessibility
A knowledge node can be activated either when a person is confronted with external information or when internal information is retrieved from long-term memory (Keller 1993). The fundamental requirement to activate of any brand association is that consumers are aware of the brand. Brand awareness reflects consumer’s ability to identify the brand under different conditions and is related to the strength of the brand node in memory (Keller, 1993). Brand awareness consists of brand recognition (consumer’s ability to confirm prior exposure to the brand) and brand recall (consumer’s ability to retrieve the brand from memory when given the product category as a cue) (Keller, 1993). Even when brand awareness exists and a link between an attribute and a brand is present in memory this link may not be accessible at any point in time (Romaniuk 2013). Thus, the question arises what determines which nodes or associations are activated at the time of the knowledge being sought. According to Wyer (2008) the accessibility of knowledge depends on four main criteria.
The first determinant is the strength of association between one knowledge node and the originally activated stimulus node through situational, informational or internally generated circumstances (Wyer 2008). The links between these nodes can vary in strength, depending on the distance from each other (Henderson et al. 1998). A link can be enhanced each time two events co-occur. For instance, when a benefit occurs during a brand experience, the link between the brand name and the benefit will be strengthened (Osselaer and Janiszewski 2001). Although it is possible to build an association with only one pairing, repeated pairings of two linked stimuli is the key to a associative learning process (Stuart et al. 1987). On top of that, a link to a brand in a consumer’s memory is stronger the more
consumers are exposed to brand communications or have had prior experiences with the brand (Aaker 1991).
Other determinants that influence the accessibility of knowledge are the recency and frequency with which the knowledge has been activated and used in the past. Intuitively, knowledge that is relatively “fresh” in consumer’s minds is more likely to be re-activated (Wyer 2008). The same applies to frequency. The more consumers encounter particular information it becomes chronically accessible in memory and is less dependent on situational factors in order to be activated. Initially, the effects of recently activated knowledge override the effects of frequency, however the effect of recency is short-lived and decreases over time, while the effect of frequency persists and has a greater impact in the long-run (Wyer 2008).
The last determinant that influences the accessibility of knowledge is the amount of processing a person is willing or able to undertake. The more consumers process a piece of information cognitively, the greater the number of elements associated with this information (Wyer 2008). This in turn positively influences the size of the associative network in which the knowledge is embedded. However, according to the Elaboration Likelihood model (Petty and Cacioppo 1986) a person must have the motivation and the ability to retrieve knowledge from memory and to process it cognitively. If consumers have a lack of motivation (because they don’t care about a product or service) or if they lack the ability (because they do not know much about the brand) (Keller 1993), they are more likely to have a small amount of linked associations with the original knowledge node and therefore will have a smaller associative network.
3.3. Knowledge Activation Order
The spreading activation theory implies that the order of activated associations depends on the content and cue. What linked associations get activated in hindsight depends on the relative strength of the connection (Franzen and Moriarty 2009). As mentioned before,
associative networks of brands consist of different types of associations. Previous research has shown that some types of associations or thoughts come to mind more easily than others (Kahneman 2003) and therefore have an influence on the activation order.
3.3.1. Category Associations
!
Consumers are confronted with an overwhelming amount and variety of information in their environment on a daily basis. To be able to deal with all the encountered information, people tend to group objects and events into categories based on perceived similarities. Categorization is considered as a fundamental cognitive activity (Mervis and Rosch 1981), which allows consumers to structure and simplify information in memory (Rosch, 1975) as well as to enhance information processing efficiencies (Bruner, Goodnow and Austin 1956). In buying situations, consumers categorize products in order to identify and evaluate product-related information (Cohen and Basu, 1987). They develop a categorical structure of brands and their associated attributes in memory (Punj and Moon 2002), that has an influence on how consumers evaluate brand and attribute level information (Collins and Loftus 1975). Keller (2013) emphasizes that a product category structure exists in a consumer’s minds with product class information at the highest level and product type and brand information at the lower level. According to Nedungadi (1990), product category associations are often activated first in consumers purchase considerations. This category knowledge serves as a basis for later comparison with other brands and for consumer judgments (Cohen and Basu, 1987).
3.3.2. Performance Associations
!
When product category knowledge is activated, consumers are likely to transfer this information to brands and products within this category. The focus and attention is then guided to relevant product features and characteristics (Murphy and Ross 1994). Olsen and Muderrisoglu (1979) conducted a study where a free elicitation procedure was used to
examine the associations attached to products from three different product categories (toothpaste, ballpoint pens and blue jeans). For all product classes, approximately 70% of the elicited associations were related to product attributes or characteristics. Timmerman (2001), who by using a free association method investigated the core associations of twelve different brands in different product categories in the Netherlands, found similar results. Product-related associations such as product attributes and price were found in nearly all of the brands in the study. This evidence gives reason to suggest that associations related to the performance of a product are activated relatively easy.
3.3.3. Imagery Associations
!
To recap, brand imagery associations reflect more intangible aspects and how consumers think about a brand abstractly, instead of what the brand actually does. Dukes and Bastian (1966) conducted an experiment to test whether abstract or concrete words are recalled more frequently. The results showed that words representing something concrete (such as “friend”, “island”, “picture” or “vinegar”) were recalled slightly better than words representing an “abstract” meaning (such as “century”, “boredom”, “hatred”, or “health”). Accordingly, it can be argued that it is harder for consumers to retrieve abstract associations related to brand imagery (such as brand personality or values, for example) compared to concrete associations related to brand performance.
In contrast, the Elaboration Likelihood Model (Petty and Cacioppo 1981) suggests that consumers not always engage in extensive cognitive processing and careful consideration of product attributes in order to form an attitude towards a brand. In low-involvement conditions, in which consumers lack the motivation or the ability to judge product-related information, consumers often rely on peripheral cues. These cues can include celebrity endorsers or the country-of-origin, the store where the product is sold, or situations in which the product can be used. Applied to the association elicitation context, it seems reasonable to
suggest, that consumers who either don’t care about or don’t feel like they possess product-related knowledge, will recall more peripheral, imagery-product-related associations. This is in line with Aaker (1991) who argued that it is easier for consumers to retrieve intangible information from memory compared to detailed product attributes.
3.3.4. Judgements
When making judgments about a product or service, consumers either rely on attitude-based or attribute-based evaluations (Czellar and Luna 2010). Attribute-based processing represents the process when consumers take multiple concrete product attributes into consideration to evaluate a product and form an overall judgment. This process requires cognitive effort from the consumer since these attributes have to be retrieved from memory before making a judgment. The evaluation based on global attitudes on the other hand does not require much cognitive effort. Behavioural (Bargh, 1997; Zajonc, 1998) as well as neuropsychological science (e.g., LeDoux, 2000) lends evidence, that the assessment of whether an object is good or bad is carried out quickly and efficiently. Depending on whether consumers form an attitude or attribute-based evaluation, these thoughts come to mind with varying ease.
3.3.5. Feelings
According to Fiske and Taylor (1995), brand associations can be stored in consumers’ memory in form of emotional impressions. Brand emotions or feelings are neural, non-verbal assessments of experiences with a brand. For example, the Coca Cola advertising in which a handsome young man drinks a can of Coca Cola, can trigger emotional reactions stored in a consumer’s memory. These feelings can dominate the decision-making in low involvement conditions where consumers do not spend a lot of time and cognitive effort to take alternatives into consideration (Supphellen 2000). Even though these feelings are stored in the associative network, they are often unconscious (Plutchik 1993). For this reason, these brand associations are harder to access and retrieve from memory (Supphellen 2000).
Another problem is that consumers have difficulties to verbalise unconscious associations such as feelings. To verbalise how consumers feel about a brand requires active cognitive processing and not everybody is willing to undertake this effort (Supphellen 2000).
3.3.6. Resonance
Some consumers can develop an intensive relationship with a brand because they have the feeling that they are “in sync” with it (Keller 2013). These consumers are very brand loyal and oftentimes engage in brand communities. According to Muniz and O’Guinn (2001, p. 412) “a brand community is a specialized, non-geographically bound community, based on a
structured set of social relations among admirers of a brand.”
Classic examples of brands, which have achieved brand resonance are for instance Harley Davidson or Apple (Keller 2013). Some practitioners (Roberts 2004) and academics (Batra et al 2012) even go so far to say that some consumers “love” a brand. According to Roberts (2004) a brand has to appeal to the head and the heart of consumers. This suggests, that it might be easier for symbolic brands to achieve brand resonance than for functional brands, which usually only appeal to the head. Batra et al (2012) argue that an important element of brand love is that consumers have frequent thoughts about the brand. But how often will thoughts about this deep psychological bond be elicited when consumers are asked about a brand? Batra et al (2012) described that participants in their study had difficulties to describe brand love, since it is tacit knowledge and therefore not easily to verbalize. Hence, associations related to brand resonance are expected to be caused less frequently.
4. Factors influencing the Size and the Content of Associative Networks
It is widely accepted that expertise, brand familiarity, involvement and other factors affect consumers’ interrelated cognitive structures for a particular concept (Alba and Hutchinson 1987; Sujan 1985). As a consequence, consumers differ in the number of associations stored in memory for a brand, the content of these associations and their structure. Another factor which can influence the elicited content and structure of the associative networks is its size.
4.1. Consumer Expertise
Expertise is defined as “cognitive competence” in a domain (Sternberg and Frensch 1992, p. 191) and is considered to be more abundant, elaborate and efficiently organized than the knowledge of novices (Czellar and Luna 2010).
The degree of knowledge has an influence on the availability and accessibility of information and in turn also on the size and the content of associative networks. While availability describes whether or not information is stored in consumers’ long-term memory, accessibility refers to the potential of this information being activated. In contrast to novices, experts have a greater availability of detailed attribute-related information about a specific domain (Czellar and Luna 2010). The product or category knowledge stored in their memory, is more accessible. Ergo, the size of the associative networks of experts is likely to be larger than the associative network of novices.
However, the degree of knowledge not only influences the size, but at the same time also affects the content of the associative network. Previous research has shown evidence that experts base their brand evaluations on multiple specific and concrete product attributes while novices rely on more global attitudes like first impressions or stereotypes (Dillon et al. 2001) to evaluate a brand. Czellar and Luna (2010) give an example of an expert and a novice about BMW. The authors argue that when novices form an opinion about the brand,
they would rely mostly on general stereotypes such as “expensive” or “European” and can only elicit a few specific product attributes. In comparison, experts activate multiple concrete product attributes about BMW such as the “specific engine”, “fuel efficiency” or “design aspects” in order to evaluate the brand.
Applied to a free association method, it seems reasonable to suggest that experts would mention more functional attributes such as specific product attributes, rational benefits and information about the product class, while novices may mention more non-functional attributes such as intangibles, psychological benefits, celebrity endorsers or the user group of a brand. For example, car experts are also more likely to mention more judgements or feelings about different car brand because they have been expressing their attitudes many times in different contexts, while it may be more difficult for novices to form and express their attitudes towards a brand (Czellar and Luna 2010).
4.2. Brand usage and familiarity
An important source of brand knowledge is personal experience (Alba and Hutchinson 1987). Associations are developed based on numerous exposures to and experiences with the products carrying a certain brand name and then being transformed into associations of a brand as a whole (Oakenfull and McCarthy 2010). As familiarity with a brand increases, the cognitive structures used to differentiate products become more refined and complete (Alba and Hutchinson 1987). The authors suggest that this is due to the improved ability to elaborate on specific information and to remember product information. Brand familiarity can be conceived as the number of product-related experiences that have been accumulated by consumers through advertising exposure, information search, interactions with salespersons, purchase decisions or product usage (Oakenfull and McCarthy 2010).
Fazio and Zanna (1981) conducted a study in consumer psychology and found that direct experience with a brand has the greatest impact on a consumer’s associative network.
Consistent with the findings of Alba and Hutchinson, Oakenfull and McCarthy (2010) proposed that consumers who have more experience with a brand tend to be more familiar with the brand and therefore have a more multi-dimensional associative network (Low and Lamb 2000). Since customers of the brand have a more extensive knowledge structure, they are likely to recall a greater total number of brand associations from memory compared to non-customers (Oakenfull and McCarthy 2010).
Given that the increased usage results in a more complex, but also more accurate brand knowledge structure, the type of brand associations, which are held in memory are likely to vary from light to heavy users. Moreover, experiences with a brand are shown to reduce the cognitive effort during decision-making (Hoyer 1984). As consumers only possess a limited amount of processing capacity, those who have more experiences with a brand are more likely to recall a greater number of brand specific associations (Oakenfull and McCarthy 2010). Brand-specific associations are those attributes or benefits which are unique to a brand and differentiate it from competitors. They can be referred to as points-of-differences (Keller et al. 2002). Brand-specific associations reflect either product-related attributes or imagery-related associations such as brand personality for instance. These associations provide more information and therefore a greater cognitive effort of processing (Oakenfull and McCarthy 2010). Contrarily, consumers with less experience with a brand will not have as many brand specific associations. Instead, they rely on a more limited memory structure, consisting of product-category associations. Product-category associations are those associations not unique to a brand, but rather shared with other brands in the same product category and therefore can be considered points-of-parity (Keller et al. 2002).
4.3. Purchase intension, category involvement and attitude
The elaboration Likelihood Model (Petty and Cacioppo’s 1986) argues that consumers are more willing to process information having a greater personal relevance. A consumer who is
about to purchase a product is likely to consider multiple aspects such as product attributes of brand x and its competitors, his preferences about the design or he might recall prior experiences with the brand. During these purchase considerations and decision-making situations consumers have a higher level of involvement. This leads to the activation of knowledge stored in the in consumers’ minds and can evoke processing of more external information (Franzen and Moriarty 2009). The same is true for consumers who have a positive attitude towards a brand. They are more likely to engage in cognitive processing and a more extensive, larger associative network will be elicited.
4.4. Size of Associative Networks
The size of the associative network of a brand affects the memory structure in consumers’ minds (Krishnan 1996) and might have an influence on the elicited content of associations.
As the number of brand associations increases, the structure of the associative networks becomes more complex, which makes it generally easier to access a particular node through multiple pathways for retrieval (Krishnan 1996). Because of these multiple pathways and connections between the memory nodes, various types of associations get activated.
However, at the same time, as the number of associations increases the structure of the associative network becomes more complex. This leads to lower memory of a particular association due to interference with other linked associations (Meyers-Levy, 1989). As a result, the likelihood that a particular node is activated decreases since the number of competing association increases. The more knowledge nodes are activated, the less intensively each one will be activated. This phenomenon is called the “fan effect” (Anderson 1983a, 1983b). The interference of associations is likely to be weaker for mature brands in comparison to new brands since mature brands had time to develop strong associations in consumers’ minds (Kent and Allan 1994).!
5. Hypotheses Development
In the previous chapter it was argued that the size of associative networks influences the knowledge structure and information retrieval processes. When the size of associative networks gets larger its structure becomes more complex since different knowledge nodes are connected with each other. Due to the spreading activation process from knowledge node to other linked nodes, various associations can get activated. With an increase in the number of associations the knowledge stored in consumer’s minds gets more interrelated and therefore it becomes easier to access. Ergo, the likelihood that a consumer retrieves different categories of brand associations from memory increases with the size of the associative network. Therefore, I propose:
H1a: As the total number of brand associations increases the number of covered categories within the Brand Resonance Pyramid will increase.
Since these cognitive processes are the same for functional as well as symbolic brands, it is expected that one will find the same effect for both brand concepts. Therefore, it can be proposed:
H1b: The increase in the number of covered categories within the Brand Resonance Pyramid is not dependent on the type of brand concept.
When the number of associations increases the structure of the associative network becomes more complex, which causes a lower memory of a particular association due to interference with other associations that get activated simultaneously (Meyers-Levy, 1989). Therefore, it is less likely that a particular node is activated since the number of competing association increases.
Combining this knowledge with the evidence from past research that category associations are often activated first, it can be proposed that there will be a larger proportion of category associations within small associative networks. When the number of associations increases, these associations will get activated less intensively due to the interference with other associations. Hence, within larger associative networks, there will be a smaller proportion of category associations compared to small associative networks. Thus, the following hypothesis can be made:
H2a: As the total number of brand associations increases the proportion of associations related to the “Salience” category will decrease.
Since categorization is considered a fundamental cognitive activity in order to be able to process information as it was discussed in the previous chapter, there is no reason to expect that this effect would be different for functional and symbolic brands. As a consequence, I propose:
H2b: The decreases in the proportion of associations related to the “Salience” category will not depend on the type of brand concept.
In addition, past studies have shown that associations related to products and product characteristics were found in the majority of associative networks. Based on these findings it was argued that these associations are activated frequently and relatively easy. Therefore, it can be expected that one will find a large proportion of product associations in small associative networks. Nevertheless, when the network gets larger a smaller proportion of product-related associations will be found, since other nodes get activated and will cause interference. For this reason, the following hypothesis is introduced:
H3a: As the total number of brand associations increases the proportion of associations related to the “Performance” category will decrease.
It is generally accepted that functional brands have more functional associations related to brand performance compared to symbolic brands. Consumers buy functional brands to fulfil a functional need and therefore are more likely to consider multiple product characteristics and features. So in small associative networks functional brands have a larger proportion of associations related to brand performance compared to functional brands. When the associative network gets larger, this proportion will decrease for the functional as well as for the symbolic brand due to the interference with other associations. However, because the proportion of associations related to brand performance for the symbolic brand isn’t as large as the proportion of the functional brand in small associative networks, the decrease will be smaller for the symbolic brand compared to functional brand. Accordingly, the following hypothesis can be made:
H3b: The decrease in the proportion of associations related to the “Performance”
category will be stronger for the functional brand compared to the symbolic brand.
We have argued that individuals often rely on peripheral cues when processing information in low-involvement conditions (Petty and Cacioppo 1981). When participants are asked to mention everything that comes to their mind when they think of a brand, they are likely to run out of thoughts after a few associations. To constrain the cognitive effort while searching for more associations in memory participants are expected to retrieve more peripheral associations which are related to the imagery category. Consequently, the probability of finding a larger proportion of associations related to “Imagery” is higher for large associative networks compared to small ones. Hence, it can be proposed:
H4a: As the total number of brand associations increases the proportion of
Since associations related to “imagery” are located on the right side of the Brand Resonance Pyramid, they belong to the symbolic associations. Symbolic brands will elicit more of these symbolic associations compared to functional brands. Since consumers attach fewer symbolic associations with functional brands, they are not likely to elicit much more symbolic associations when the associative network gets larger. In contrast, with an increase in the total number of associations, symbolic brands are expected to evoke even more imagery associations. For this reason, I propose the following:
H4b: The increase in associations related to the “Imagery” category will be stronger
for the symbolic brand compared to the functional brand.
It was argued that consumers have difficulties to form an attribute-based judgment. Experts find it generally easier to make an evaluation since they probably have formed judgments about products from a product class and expressed their attitude towards these brands before. Because of these difficulties, judgements are likely to be activated later in the free elicitation procedure. When consumers make the cognitive effort to mention multiple associations the probability that they think actively about product characteristics and form a judgement on this basis about the product increases. Therefore, the following will be proposed:
H5a: As the total number of brand associations increases the proportion of
association related to the “Judgments” category will increase.
Since the rational formation of judgments about brand performance plays a more important role for functional brands compared to symbolic brands, we expect to find more judgements in large associative networks of functional brands compared to symbolic brands. For this reason, the following will be proposed:
H5b: The increase in the proportion of association related to the “Judgments”
category will be stronger for the functional brand compared to the symbolic brand.
Feelings are harder for individuals to articulate verbally and hence, are less likely to be found in small associative networks since the articulation requires cognitive effort and not everybody will have the motivation or ability to do so. Accordingly, when consumers think actively about the associations they attach to a brand and mentions more associations, then the probability that some of these associations include feelings increases. Thus, the following hypothesis can be made:
H6a: As the total number of brand associations increases the proportion of
associations related to the “Feelings” category will increase.
Because symbolic brands address consumers’ psychological and social needs, instead of the pure functional need, they evoke more emotional feelings compared to functional brands. Hence, we expect the following:
H6b: The increase in the proportion of associations related to the “Feelings”
category will be stronger for the symbolic brand compared to the functional brand.
In the previous chapter it was argued that not all brands achieve resonance. Moreover, feelings such as brand love are tacit knowledge and therefore not easily to verbalize for individuals. Since it requires cognitive effort to verbalize these feelings and not every participant will engage in active thinking, the probability to find associations related to brand resonance will increase with the size of the associative network. Ergo, the following can be proposed:
H7a: As the total number of brand associations increases the proportion of
associations related to the “Resonance” category will increase.
Brand resonance describes the psychological bond that customers have with a brand (Keller 2013). This intense relationship is more likely to exist for symbolic brands instead of
functional brands because they create an emotional connection to consumers, which is crucial to achieve brand resonance. Accordingly, the following hypothesis can be made:
H7b: The increase in the proportion of associations related to the “Resonance”
category will be stronger for the symbolic brand compared to the functional brand.
6. Data and method
6.1. Stimuli Development
The interest of this study lies in understanding whether the size of associative networks has an influence on the evoked associations of functional and symbolic brands. Sportswear and sports shoes were selected as a product category to test the posted hypotheses. This product class was chosen since it fulfils several requirements.
First, the selected category contains a number of brands with a symbolic brand concept and brands with a functional brand concept. Bhat and Reddy (1998) chose sports shoes as one of the product classes to examine the brand concept with Nike as a symbolic and Converse as a functional brand.
Second, the product class is commonly used (Bhat and Reddy 1998) and consumers generally have sufficient knowledge of the main brands of sportswear and sports shoes, since it is likely that they have tried more than one brand (Belen Del Rio, Vasquez, and Iglesias 2001). This is in line with Leuthesser et al. (1995) who recommended analysing brands which are sufficiently well-known by consumers.
Third, the sportswear and sports shoes category contains mature brands such as Nike and Adidas which had time to establish multiple brand associations in the minds of consumers. This is in alignment with Krishnan (1996) who selected two mature brands for each product category to investigate the characteristics of memory associations.
Lastly, since sportswear and sports shoes are visible consumer products, they have some technical aspects. Also brands in this category invest a lot in advertising and sponsorship which explains why consumers are likely to attach various types of associations to the brands. For example, associations about the user imagery, product features or celebrity endorsers can be evoked.
To select brands which are sufficiently well-known by consumers a preliminary study was conducted. It tests brand recall when given the product category sportswear and sports shoes as a cue.
As shown in the table below a list of 40 different brands was obtained. The category leaders Nike and Adidas were mentioned by every of the 25 respondents, followed by Puma, Reebok and Asics which were mentioned by more than 37,5% of the participants. The number of people who mentioned one of the remaining brands is drastically lower.
Table 2: Pre-Test 1 (Brand Recall)
Brand name Number of mentions Brand name Number of mentions
1. Nike 25 21. Fjall Raven 1
2. Adidas 25 22. Jack Wolfskin 1
3. Puma 20 23. Kappa 1
4. Reebok 18 24. Lining 1
5. Asics 15 25. Lonsdale 1
6. New Balance 5 26. Moncler 1
7. Champion 3 27. Meindl 1
8. Helly Hansen 3 28. Mammut 1
9. Mizuno 3 29. Moncler 1
10. Under Amour 3 30. Peak 1
11. ODLO 2 31. Phenix 1
12. Kempa 2 32. Roxy 1
13. Saucony 2 33. Salomon 1
14. Le coq sportif 2 34. Spalding 1
15. And1 1 35. Schöffel 1
16. Björn Borg 1 36. Umbro 1
17. Brooks 1 37. Venice beach 1
18. Burton 1 38. Vibram 1
19. Converse 1 39. Warrior 1
Based on these results a subset of brands was chosen for a second pre-test. The purpose was to identify which brands are perceived as either more functional or more symbolic. This is crucial for the intended manipulation in the main study. Nike, Adidas, Puma and Reebok were thought of as possible symbolic brands whereas New Balance, Asics, Brooks and Mizuno were thought of as possible functional brands. Following Bhat and Reddy (1998), a seven-point Liket scale to measures a brand’s functional or symbolic value was used. Twenty respondents were asked to rate their agreement with twelve statements, which were selected from the list of eighteen questions about the brand concept adopted from Bhat and Reddy (1998). For example, brand functionality was evaluated by asking participants to rate their agreement with the statement “the brand is for people who are down-to-earth” from (1) strongly disagree to (7) strongly agree. Similarly, with respect to brand symbolism, participants were asked among other things to rate whether “people use the brand as a way to
expressing their personality”.
For each brand, the ratings of the respondents on each question were summed up and divided by the number of participants. Afterwards, these scores were added to the scores for the other functional/symbolic questions and divided by the number of questions. This resulted in a mean functional and a mean symbolic score for each brand, as it is depicted below.
Table 3: Pre-Test 2 (Brand Concept)
Brand Functional Symbolic
Nike 4,56 (1,44) 6,13 (0,65) Adidas 4,93 (1,19) 5,48 (0,88) Puma 5,14 (1,06) 4,50 (1,31) New Balance 4,89 (1,11) 4,65 (1,51) Reebok 5,32 (1,07) 3,55 (1,40) Asics 5,79 (1,00) 3,75 (1,27)
