Revitalization or devitalization of a brand through a step-up brand extension ? What is the right distance?

R

EVITALIZATION OR DEVITALIZATION OF A

BRAND

THROUGH

A

STEP

-

UP

BRAND

EXTENSION

?

W

HAT IS THE RIGHT DISTANCE

?

M

MSC IN BUSINESS ADMINISTRATION

– MARKETING

S

TATEMENT OF ORIGINALITY

This document is written by Student Trix Klerks who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

T

ABLE OF

C

ONTENTS

Horizontal brand extension ... 10

Vertical brand extension ... 10

Step-down brand extension ... 10

Step-up brand extension ... 11

Positive effects ... 12

Negative effects... 13

Effect of distance on parent brand evaluation ... 15

The influence of brand ownership ... 15

... 17

METHODOLOGY ... 18

Pre-test ... 18

Experimental Design ... 19

Participants ... 20

Procedure and measurements ... 20

RESULTS... 23

Preparing analyses ... 23

Descriptive of the total sample... 23

Descriptive of different groups ... 24

Correlations ... 27

The effect of gradual increase in step-up brand extensions on parent brand evaluation ... 29

Additional analysis: the effect of perceived quality of the step-up brand extension on parent brand evaluation... 30

The effect of brand ownership on the relationship of gradual increase in step-up brand extensions on parent brand evaluation ... 32

Additional analyses ... 34

DISCUSSION ... 38

The effect of step-up brand extensions on parent brand evaluation ... 38

Brand ownership ... 41

Limitations and recommendations ... 42

Managerial implications ... 44 CONCLUSION ... 44 REFERENCES: ... 46 APPENDICES ... 48 Appendix A ... 48 Appendix B ... 54

A

BSTRACT

Managers often extends brands to a higher level of quality and price with the aim to increase sales and improve profit margins, however these step-up brand extensions have an effect on consumer’s evaluation of the parent brand. Prior research on step-up brand extensions produced inconsistent results and has been based on various theories, most notably the Association Network Theory and the Bookkeeping model which have led to different predictions. Drawing on these theories, current research expected to find different results for different price and quality levels of step-up brand extensions. It was expected to find a positive effect of a small step-up extension on the evaluation of the parent brand, and when a step-up extension is too far from the parent brand in terms of quality and price this effect would into a negative evaluation. Results in this study showed no effect of step-up brand extensions on the parent brand evaluation. This study showed that quality perceptions are not necessarily equal among all consumers. However when the perceived quality of a step-up brand extension increases this will lead to an increase in parent brand evaluation. From results in current study can be concluded that the general assumption that a step-up brand extension that is too deviating will by definition lead to a negative parent brand evaluation should be reconsidered. There is no consensus yet about the conditions under which negative effects will occur. Interestingly, affective commitment seems to affect the relationship of step-up brand extensions on parent brand evaluations. Further research could examine the effect of affective commitment on different step-up brand extensions in different product categories.

I

NTRODUCTION

What does a consumer think of when he or she thinks of Volvo? Most consumers have associations like: safety; reliability; status; family man; Sweden and so on. Volvo’s brand name is a strong name with strong associations. For Volvo this brand name and its accompanied associations are of extremely high importance. This brand name eliminates a huge risk for Volvo of the success of the sales of their cars. However, what if a certain brand name has lost some of these positive associations? Rebuilding these associations by reframing or repositioning its entire image is an extremely complicated and high-risk operation. What if there was a safer and less complicated way to improve a core brand image, without the big risk of changing and ruining it? What if a brand image could be enhanced with the help of the introduction of a new product under that same name?

Strong brand names are of incomparable value. Truly strong brand names considerably enlarge the success of the introduction of a new product under that name. Therefore, it is very tempting to “extend” and exploit a brand name. During the past decade brand extensions have been the driver of strategic growth of a variety of firms (Aaker, 1990). Brand extensions are usually introduced to leverage the equity of an existing brand and introduce a new product relatively easily (Völckner & Sattler, 2006). A second reason for introduction of a brand extension is to affect the brand image positively (Balachander & Ghose, 2003). In order to reach those goals, brands get extended in different ways. Horizontally, through new services or products, to touch new product markets. And vertically, extensions are introduced in the same product line in order to hold or increase revenues by targeting a broader audience or focus on a niche segment who are willing to pay higher prices (Goetz, Fassnacht & Rumpf, 2014). This means similar products in different quality and price ranges, to attract customers with different purchasing powers. Vertical extension come in two ways. Firstly, step-down extensions, the extension of a brand to a lower priced level, can be enormously profitable and are meant to promise increased volume and economies of scale (Aaker & Keller, 1990). Secondly, successful step-up extensions, extending a brand to a higher priced level, have relatively high profit margins while only a low increase in costs (Aaker & Keller, 1990). In this study step-up brand extensions will be evaluated, hence a successful upward extension could be of

great value to a firm. Linked to the need of revitalization of a brand name, research by Munthree, Bick and Abratt (2006) suggests, that a step-up brand extension even could have the value of revitalizing the core brand of a firm.

However, in the field of brand extensions there are conflicting theories about the possible outcomes of step-up brand extensions. On the one hand there are findings that support the positive consequences of a step-up brand extension, on the other hand there are findings for negative consequences. Based on the Association Network Theory, multiple studies found that that parent brand equity is evaluated more favorably after an extension to higher quality levels in the same product category, due to positive associations of the brand extension which spillover to the parent brand (Lei, De Ruyter & Wetzels 2008; Randall, Ulrich & Reibstein, 1998; Heath, DelVecchio & McCarthy 2011). In contrast with these positive findings, other studies have found negative effects of a step-up extension on the parent brand evaluation. Kim and Lavack (2001) found that the introduction of a step-up extension will lead to negative effects, because the extension differs too substantially from the parent brand. These findings are in congruence with the Bookkeeping model of the Categorization Model (Weber & Crocker, 1983). According to this theory, the difference in quality and price between two products in the same product category are too confusing which will lead to a negative parent brand evaluation (Kim & Lavack, 1996, Riley, Pina & Bravo, 2013). Consequently, there are two areas with contradicting convictions about the effects of step-up brand extensions.

Because of the different existing theories, further examination of the step-up brand extension field is needed. Which kind of step-up brand extensions lead to a positive effect on the parent brand and for which kind of step-up extension does a negative effect arise? Specifically, two theories are at heart in line with each other, but have conflicting conclusions. The Association Network Theory claims that high quality associations of the brand extension transfer to the parent brand, while the Bookkeeping model claims that these differences in associations between the parent brand and the extension lead to confusion and therefore to parent brand dilution. One would expect that when the difference between the parent brand and a step-up extension in terms of price and quality is larger,

has to be established when the Association Network Transfer theory is in effect and when the Bookkeeping model is in effect. Consequently, for this research the following question arises: does the distance between the parent brand and the step-up brand extension explain when positive association transfer takes place and when confusion and therefore parent brand dilution arises? Deviation in step-up brand extensions will arise because of increasing distance from the core brand regarding price and quality. In this research therefore will be investigated what the maximum distance of price and quality is between the core brand and extension to lead to a positive association transfer.

Furthermore, the existence of inconsistent findings in theory suggest that there is a factor that changes the nature of the relationship between step-up brand extensions and parent brand evaluations. According to Kirmani, Sood and Bridges (1999) consumers interpret discrepancy between the parent brand and the extension differently. Research by Kirmani et al. (1999) has shown that brand owners as opposed to non-owners, respond more favorably to the introduction of a step-up brand extension. Being more positive towards the extension is likely to have an effect on the feedback effect on the parent brand (Kirmani et al., 1999). In this research I will examine whether brand owners are more lenient regarding the distance between the step-up extension and the parent brand.

Answering this question is important for the marketing field since the introduction of a step-up brand extension could be a great solution for brands in need of revitalization operating in a saturated market (Munthree et al., 2006). Revitalizing a brand through a line extension offers a relatively low-risk opportunity to boost a brand without immediate change of the whole parent brand (Aaker, 1990; (Sullivan, 1990). By researching to which distance a step-up brand extension has potential positive effect on the parent brand and from which distance on a step-up extension could harm a parent brand, provides a useful guide for marketing practice. Theoretically lots of research has been done on brand extensions, only little have been done on step-up brand extensions. Within the research that has been done there is still lots of debate on the effect of a step-up brand extension. This research offers new insights in underlying explanations and will provide circumstances in which a step-up brand extension will lead to a more positive or negative evaluation of the parent brand. The findings of this research also offer new insights in the possibilities for brands and the condition a brand has to hold in order to seize those opportunities.

The following structure can be expected from this research. Firstly, the study provides a review of the literature relevant in the field. This review will lead to better understanding of the different concepts, theories and perspectives. Secondly, the research design and methods will be explained. Thirdly, the results of the experiment will be presented. Fourthly, the results of the research will be evaluated and there will be an evaluation of the implication of this research for theory and practices, including suggestions for future research. Finally, core findings will be discussed and conclusions will be drawn.

L

ITERATURE

R

EVIEW

Brand extensions

Firstly, I will discuss what a brand extension is, what its underlying drivers are, and what the possible consequences of brand extensions are.

As abovementioned, strong brand names are of incomparable value, because truly strong brand names considerably enlarge the success of the introduction of a product under that name (Aaker, 1990). The introduction of a new product under a similar brand name is called a brand extension. For most brand extensions, an established brand name is used to introduce a product and exploit the established brand equity for a successful introduction of the new product (Kim & Lavack, 1996). Aaker and Keller (1990) state that the risk of failure of introducing a completely new brand is much larger than introducing a product under an already established name. The familiarity consumers already have with the existing brand name helps introduction into the market. The existing familiarity helps reaching the target audience more easily. The new product can use the current favorable associations of the core brand (Lei, De Ruyter & Wetzels 2008; Randall, Ulrich & Reibstein, 1998; Heath, DelVecchio & McCarthy 2011). Also, the risks of building a new brand are much higher than introducing products under an existing brand name (Sullivan, 1990). Existing brands take the advantage of existing brand name recognition and image. Whilst by introducing a new brand, the building of brand image, recognition and equity has to start from scratch. In today’s competitive market it is a challenge to build a new brand and building equity and unique positive associations can

take a long time. For that reason, many brands choose to introduce new products under existing brand names (Aaker, 1990).

Horizontal brand extension

Horizontal brand extensions are the introduction of new products in a different product category under a familiar brand name (Sheinin & Schmitt, 1994). Horizontal extensions can be done in different gradations, a slightly related category, or by entering a completely different category. In a horizontal brand extension, the brand equity of a brand gets exploited and is used for a product in a different category (Aaker & Keller, 1990). A profitable manner, when the extension is introduced under the right circumstances (Martinez & Pina, 2003), to make use of existing brand equity. An example of the introduction of a horizontal brand extension is Lay’s chips introducing a new flavor of chips. This is the introduction of an extension within a very similar category, a line extension. An example of the introduction of a product in a new category could be, Virgin airlines introducing a Virgin healthcare insurance. A precondition of a horizontal brand extension to be successful is the level of fit between the product classes. This level of similarity between the two products and associations is highly important in the evaluation of the brand extension (Aaker & Keller, 1990).

Vertical brand extension

A vertical brand extension, as opposed to a horizontal extension, is the brand extension in a similar product category as the core brand, only different in quality and price level. Vertical extensions are usually introduced in order to reach an audience with different purchasing powers (Kim & Lavack, 1996). Vertical extensions can go in two directions, upwards and downwards.

Step-down brand extension

A step-down extension is the extension of the core brand towards a lower quality and price. This can be a valuable strategic move, since consumers transfer the associations of high quality and price to the extension (Aaker & Keller, 1990). Brands aim to capitalize on the well-known brand equity that is

built up in the core brand (Kim & Lavack, 1996). This makes consumers feel like they have the ability to purchase a high-end product for a lower price. It is a strategic move for firms since they can attract a much bigger market by touching a new and usually bigger audience for their products. It is generally believed that the link between the core brand and the extension will help in gaining acceptance for the newly launched extension (Broniarczyk and Alba, 1994). Implementing this is interesting in times of crises when the average purchasing power is lower. Step-down extensions are also useful when there are declining sales in existing markets due to new entrants or higher competition. Another reason to do a step-down extension could be due to a large opportunity in growth in the downward market because of booming segments (Aaker, 1990). However, the introduction of step-down extensions is not without risk since the introduction of a lower quality product under the same brand can have negative consequences for the parent brand. As for example a step-down extension could lead to dilution of the core brand for luxury brands (Magnoni & Roux, 2012)

Step-up brand extension

A step-up extension, on the other hand, is the extension of the core brand towards a higher quality and price. The extension upward is usually seen as more difficult to pursue, since it is less credible that a lower quality brand all of a sudden produces higher quality products. Consumers are less likely to pay a higher price because of a higher financial risk (Lei et al., 2008). However, a successful step-up extension could be very valuable for a firm. It provides the opportunity for a brand to operate in a new market segment with high profit margins, while making use of their existing assets and distribution (Aaker & Keller, 1990). Reaching these markets provides a big opportunity for sales and profit.

When products find themselves in mature markets, step-up extensions provide opportunities for brands to go premium and differentiate themselves through higher quality and higher price. These brands usually find themselves in the maturity phase and have a hard time to differentiate from other brands in that category, operating more like commodities (Young and Rubicam, 1991). Another step further, Nijssen noted in his study that “the objective of a company for introducing a line extension may be not to generate extra sales and profits, but to simply improve its strategic positioning by

introducing an improved version of its current product (e.g. with a more functional type of packaging)” (1999, p. 454).

Positive effects

Past research has found positive as well as negative step-up extensions on the parent brand. Various theories can help explain the existence of both positive or negative effects of step-up extensions on the parent brand, namely the Association Network Theory, The Social Comparison Theory, and The Categorization Model.

Firstly, The Association Network Theory explains a positive effect of a step-up brand extension on the parent brand by the possibility that the association of higher quality and price of an extension gets transferred to the parent brand. The positive associations of the brand extension spillover to the parent brand (Lei, De Ruyter & Wetzels 2008; Randall, Ulrich & Reibstein, 1998; Heath, DelVecchio & McCarthy 2011). The Association Network Theory explains that every concept has different associations connected to it in a person’s memory. For example, depending on the strength of the association in a person’s memory, when someone thinks of a rabbit he could automatically think of a carrot. Similarly, these networks of associations exist in consumer’s memory for certain brands and products. For example, many consumers will associate Volvo cars with safety. The model also explains the ease of transfer of associations from brands to different brand and product categories (Aaker, 1990). For example, because of this strong network of associations, most likely every car that Volvo produces will have the association of safety. The stronger this association of the brand is, the easier it gets transferred to a new product. The strength of an association can for example be built through repetition. Repeated exposure of pairing the association with the brand will strengthen this association (Till, 1998).

This network of associations is used in theories of brand extensions. With vertical brand extensions the brand’s associations are transferred to the extension, even though it has a different price and quality range (Aaker, 1990). However, association transfer is not a one-way process. The introduction of an extension can also change the associations that hold for the parent brand, what is

called a feedback effect (Dwivedi, Merrilees, Sweeney, 2009; Riley et. Al, 2013). The network of associations namely works in both ways, when someone thinks of a carrot he often also thinks of a rabbit. For a step-down extension, the feedback effect could be harmful when the extension for example decreases the premium association of the parent brand, which is called brand dilution. On the other hand, this feedback effect also creates opportunities. In case of a step-up extension, companies could signal greater brand expertise through this step-up extension. This feedback effect from the higher quality association to the parent brand could most favorably even lead to a boost, or revitalization of the image of the parent brand (Munthree, Bick & Abratt, 2006). According to Heath, DelVecchio and McCarthy (2011) higher-quality extensions improve the perceptions of the parent brand prestige, expertise and innovativeness. In line with the Fishbein theory a more favorable attitude towards a brand will have positive consequences on future (purchase) behavior by consumers (Fishbein & Azjen, 1975).

Negative effects

Other researchers have found a negative effect of step-up brand extensions on the evaluation of the parent brand. This negative effect can be explained by the Social Comparison Theory and the Bookkeeping model of Categorization Theory.

Firstly, the negative effect of a step-up line extension on the parent brand can be explained by the Social Comparison Theory (SCT). This theory claims that people evaluate themselves more negatively because people always compare themselves upward with more successful people. This could work similarly for brands, where the parent brand gets evaluated more negatively due to comparison with a higher quality brand extension (Goetz, Fassnacht & Rumpf, 2014). A sort of contrast effect arises, where the lower quality of the parent brand becomes more salient due to comparison with its new high-quality extension. Hence the image of the parent brand is perceived as more negative than when such a comparison is not available (Goetz et al., 2014).

The Bookkeeping model of the Categorization Model can also be used to explain why a step-up brand extension will lead to a negative effect on the parent brand. This theory describes the way in

which people create and update stereotypes about other people. The theory suggests that every piece of information a person receives about a person, incrementally updates the features of the stereotype (Weber & Crocker, 1983). It is believed that a similar process works for brands and brand extensions. Consumers hold a specific schema in their mind about the brand, i.e. the brand image. The information consumers receive about a step-up extension causes them to update their schema and so their beliefs of the parent brand, since it is incoherent with existing beliefs. A step-up extension deviates from the core brand on price and quality, this leads consumers to reevaluate the core brand. Updating the core brand does not predict the valence of the updating. Dependent on the information of the extension, the core brand will be reevaluated positively or negatively. However, according to the theory exceedingly incongruent information about price and quality may result in confusion about the image of the core brand, damaging existing beliefs and therefore leads to dilution of the core brand. Regardless of the positive or negative valence of the brand extension, merely the level of incongruity will lead to a negative evaluation of the core brand. In accordance the deviating information of the extension will lead to confusion and therefore to a less positive evaluation of the core brand (Kim & Lavack, 1996; Riley, Pina & Bravo, 2013).

Placing the Bookkeeping Model and Association Network Theory opposite of each other, one sees that both theories support the adjustment of associations of the core brand through the new information of the step-up brand extension. However, according to the Bookkeeping Model, when information of the brand extension is exceedingly deviating from the core brand, consumers get confused which leads to a more negative core brand image (Kim & Lavack, 1996). Placing these theories next to each other raises a few questions: from which point on is information of the step-up extension too deviating? To which distance of high quality and price does a step-up brand extension lead to positive association transfer to the core brand, and from which distance on is such an extension too deviating from the parent brand and therefore leads to confusion and eventually to brand dilution?

Effect of distance on parent brand evaluation

This research will analyze the condition under which a positive association transfer will take place and, under which conditions the deviation between parent brand and its step-up extension is too big which will lead to parent brand dilution.

According to the Bookkeeping Model, the discrepancy between the parent brand and the brand extension will lead to confusion and damage of the core brand image (Kim & Lavack, 1996). This research will examine how large the distance and so the discrepancy has to be in order to be damaging to the parent brand. At the same time will be analyzed to which point the difference in quality and price has a positive effect on the parent brand. To find the turning point of a positive feedback effect into a negative feedback effect, there needs to be sufficient variation in the distance of price and quality of the step-up brand extension.

Based on the Association Network Theory and Bookkeeping Model it is expected that an inverted U-shape relationship exists between the distance of the step-up brand extension and the core brand and the positive brand evaluation of the core brand. A small and moderate step-up brand extension is expected to lead to a positive evaluation, which becomes negative as distance increases. This means, a step-up brand extension has a positive effect on the core brand to the point the distance between the extension and core brand gets too confusing and therefore leads to a more negative evaluation.

H1: The effect of a step-up brand extension on the parent brand evaluation is positive for a small and moderate distance between the product and the extension, whereas the effect becomes negative when the distance between parent brand and step-up brand extension becomes (too) big.

The influence of brand ownership

I expect that as distance between the extension and the parent brand increases, inconsistency and discrepancy will overrule the positive association transfer from the brand extension to the parent brand. The discrepancy between the extension and the parent brand can be interpreted differently by

different consumers. According to Kirmani, Sood and Bridges (1999) owners of brands are more likely to respond favorably to a brand extension than non-owners. Ownership is a collection of different psychological dimensions that affect consumer’s appreciation of a product or brand, amongst which: involvement, knowledge and familiarity (Kirmani et. Al, 1999). The dimensions altogether lead to a more favorable attitude toward the parent brand. This phenomenon is known as the “ownership effect”. This theory, however, also claims that owners respond less favorably to extensions that decrease their favorite component of a brand. One could say that owners are more extreme in their evaluations of the decisions of the brand because they are more involved and knowledgeable. Owners will embrace a positive development of the brand but will be more critical towards a negative development of the brand. Research by Kirmani et. Al (1999) claims that a step-up brand extension adds value to the core brand image since it is a signal of higher prestige. Hence, a step-up brand extension is perceived as positive by brand owners, and therefore I expect that brand owners respond more positively to a step-up brand extension than non-owners.

If brand owners respond more positively to the extension, it is more likely that owners have a more positive feedback effect as well. When the ease of positive association transfer towards the brand extension is bigger, it is probable that positive associations of the step-up brand extension transfer more easily to the core brand as well (Aaker, 1990). The ease of positive association transfer is therefore expected to take place for further distances before confusion and dilution appears. It is expected that owners have a positive core brand evaluation for a bigger distance between core brand and brand extension, where for non-owners confusion and dilution already appears.

H.2: The negative effect of the step-up brand extension on the parent brand arises at a smaller distance between the parent brand and the step-up extension for non-brand owners than for brand owners.

The previous discussion is summarized in the conceptual model shown in Figure 1. The dependent variable of this study was the consumers evaluation of the parent brand (Puma). The independent variables were the different levels of increase in price and quality of step-up brand

extensions, and the moderating variable was the state of being an owner of the brand. The different step-up brand extensions involved different price ranges of 25%, 50%, 200% and 500%. Also, the quality of the extensions increased gradually.

Figure 1. Conceptual model with expected relationship between the distance between the step-up brand extension and the core brand on the evaluation of the core brand. With the moderating effect of brand ownership on this relationship

M

ETHODOLOGY

In order to test the hypotheses, this research analyzed existing parent brands with fictitious step-up brand extensions. The choice for an existing parent brand was based on extant literature on vertical brand extensions (Kirmani et al. 1999; Riley et al., 2013 etc.). Existing brands have existing knowledge, evaluations and attitudes, which are needed to test whether evaluations and attitudes can be enhanced. In order to test whether brand equity can be leveraged, there needs to be existing brand equity in the first place. Furthermore, the choice for existing brands improves ecological brand equity (Washburn, Till & Priluck, 2000). On the other hand, the choice for fictitious brand extensions is based on improving manipulation of the experiment (Kirmani et al., 1999), ensures consistent levels of brand knowledge among respondents (Kim et al., 2001), and dismisses the effect of pre-existing attitudes (Kim et al., 2001).

Pre-test

In order to select the right brand for use in the experiment, a pre-test was conducted. Based on previous studies the brand needs to meet a few demands. Firstly, the brand needed to be internationally known and have products for both men and women. The brand needed to have an average liking of the brand, since there needed to be room to measure an increase and a decrease of the brand’s liking. Consequently pricing of the brand should be average as well, in order not to have a ceiling effect, there needs to be room for increase in price for the brand extensions (Kim, Chung, Lavack & Smith, 2001). Familiarity of the brand should be high, otherwise respondents had no existing attitude towards the brand to be measured. Most literature is based on fashion, watches and cars therefore it was recommended to test in the same industry to facilitate a comparison with previous studies. There needed to be a 50/50 brand ownership in order to make a distinction between brand owners and non-owners. Which made not every brand suitable for the intended sample to address.

Category 1, sunglasses brands, contained: Ray Ban, Ace and Tate, Oakley, Prada Eyewear, Persol, Dior and Polette. Category 2, watch brands, contained: Casio, Daniel, Wellington, Seiko,

Oozoo, Zinzi, Michael Kors and Guess. Category 3, shoe brands, contained: Vans, Converse All Stars, New Balance, Adidas, Puma, Nike and Reebok. And category 4, fashion brands, contained: Topshop, Mango, Vero Moda, Diesel, Levi’s, Replay, Tommy Hilfiger and Only.

The study required the selection of one of the abovementioned brands. The sample existed of 30 participants (26 women and 4 men) who were all between 21 and 59 years old (mean age = 27). Respondents assessed familiarity (FAM), attitude towards brand (ATT), perceived value of the brand (VAL), purchase intentions (PUR) through 7-points scales (not familiar/very familiar, not likeable/likeable, not appealing/appealing, bad buy/good buy, bad value/good value, extremely unlikely/extremely likely).

From the pre-test the footwear brand Puma met criteria for the study, since it is international, has shoes for men and women, scored above the median (4) of familiarity (FAM) and scored around the median for attitude towards brand (ATT), perceived value (VAL) and purchase intentions (PUR).

Average scores of all brands on different factors were evaluated and Puma scored relatively seen best on the selected variables. It scores high on familiarity (FAM = 6,1). And around the median for attitude, perceived value, purchase intentions and affective commitment (ATT = 4,88; VAL = 5,15; PUR = 4,37; COM = 4,11). Puma scored relatively too high on perceived value, however there is still room to improve and in the footwear category there is still much room to increase in price. Puma scored on average most favorably on the different variables and is therefore selected to use in the experiment.

Experimental Design

For this study an online experiment was conducted in order to answer the research question and test the hypotheses. The online experiment was composed with use of Qualtrics. By the use of a 4 (distance of brand extensions: low (+25%), moderate (+50%), high (+200%), very high (+500%)), x 2 (brand ownership: yes vs. no) between-subjects design. Participants were randomly assigned to one of the experimental conditions. The respondents were invited to participate in the study via a link posted on social media (Facebook). In addition, half of the participants were recruited via the website

www.Profilic.ac, an international website for research respondents where each respondent got paid €0.45 for their participation.

Participants

In total 336 respondent started the experiment, of whom 320 respondents fully completed the experiment. 43.3 % of the total amount of participants was female (N = 139) and 56.6 % was male (N = 181). De average age of the participant was 27 years (M = 27.00, SD = 7.69). Most respondents (34,1%) had a Dutch nationality, second were from the United Kingdom (21,9%) and third were from the USA (8,7%).

The population of the research was represented by international consumers. This design allowed to determine the effect of gradual price and quality increase of a step-up brand extension on the evaluation of the parent brand.

Procedure and measurements

Measurement of the independent variables

There were four different questionnaires designed for this study, the four different variations were based on the four different variations in price and quality of the step-up brand extension scenario. Each questionnaire began with an introduction where privacy was reassured and gratitude for their participation was showed. For all questionnaires, the introduction contained an explanation that the survey concerned an evaluation of the Puma brand. A picture of a Puma model was showed (see Appendix B). The subsequent questions regarded brand familiarity (Milberg et al., 1997), brand ownership (yes/no), brand ownership in the past (yes/no). Apart from brand ownership, all variables were measured on a 7-point Likert scale, with higher numbers representing higher scores. Familiarity was measured with one item (not familiar/very familiar).

Thereafter the different conditions were presented in random order for different participants. In the questionnaires, respondents read the statement: Puma is considering the introduction of a new (more luxurious/very luxurious/superior luxurious) model at a price of €200 (240/320/800). It will

consist of a combination of higher (high/very high/outstanding) quality leather (and suede) and has a more comfortable foot bed (very comfortable foot bed/will have a revolutionary design leading to a high level of comfort/will have a revolutionary design leading to an outstanding level of comfort). This new model would be the first in a new line priced 25% (50%/200%/500%) above the current price range of €60 to €160”.

Measurement of the dependent variable

Subsequently, respondents were asked in a composed question to indicate on 7-points Likert scale their attitude towards the parent brand (Kirmani et al, 1999) with 2 items (not likable/very likable, unappealing /very appealing), their evaluation of the parent (Lei et. Al, 2008b) brand with 3 items (very unfavorable/very favorable, very negative/very positive, very unattractive/very attractive). Thereafter, on a 7-point Likert scale the perceived value of the parent brand was questioned (Kirmani et al., 1999) with two items (bad buy/good buy, bad value/good value).

Measurement of demographics

The last section of the questionnaire asked the respondents about their age, gender and nationality.

Measurements of covariates

The survey included a number of extra scales measuring additional variables. These variables were not included in the conceptual model, however these were included in the experiment since they could be important variables to control for or were valuable for additional analyses. Before manipulating the different conditions of the experiment, parent brand market position was measured with 2 items (budget/luxury, functional/prestige) (Milberg et al., 1997; Lei et al., 2008b).

Furthermore, after the manipulation of the experiment, as a control question the variation between the quality of the different conditions was questioned with one item (Kirmani et al, 1999; Aaker & Keller, 1990). This item questioned the evaluation of the brand extension on a 7-point Likert scale (low quality/high quality). Subsequently respondents were asked to indicate for 3 statements on

a 7-point Likert scale to what extent they feel committed to the brand Puma (Evanschitzky et al., 2006). Respondents had to indicate to what extent they: trust; identify; feel emotionally attached to the brand Puma (strongly disagree/strongly agree). This measurement was included in order to be used as an alternative for brand ownership in additional analyses.

The final questions about the parent brand regarded the consideration of the purchase intentions of Puma footwear (Aaker & Keller, 1990; Riley et al., 2013) with three items (very unlikely/very likely, I would not consider it/I would consider it, Not probably/Very probable). Purchase intentions of the parent brand was included for additional analyses as an alternative for parent brand evaluation.

Reliability of scales

Reliability checks were performed in order to examine the consistency of the scales. Below (table 1) a listed overview of the Cronbach’s Alpha of the different scales. All variables, except for market position, have a α > 0.7, which indicates that the items of a single variable are consistent in their measurement. The market position scale is included for additional analyses, so not a scale for core measurements. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above .30). Also, none of the items would substantially affect reliability if they were deleted. All scales were considered to have a good level of reliability and were therefore included in the study.

Table 1.

Cronbach’s alpha results of the various scales

Test Cronbach’s Alpha N of Items

Market position .585 2

Affective Commitment Attitude parent brand Evaluation parent brand Perceived value

Purchase Intention

Total evaluation parent brand

.761 .849 .833 .784 .945 .900 3 2 3 2 3 6

R

ESULTS

In this section a detailed description of the results of the study will be provided. All the results were analyzed using IBM SPSS Statistics 24.0

Preparing analyses

Antecedent to the analyses, the raw data has been prepared for the analysis. I checked for missing values, extreme scores, normal distribution and descriptive statistics were calculated to check for errors. There were no results deviating from the desired range, therefore I used a sample size of N = 320 in this study.

All items for parent brand evaluation, purchase intentions, market position and affective commitment were computed into a single variable representing the averages. The variables nationality, ownership and ownership past, were recoded. As for the different levels of price and quality levels. In all mentioned analyses I worked with parent brand evaluation as dependent variable and a significance level of α < .05. For following analyses accompanying assumptions were checked and no were violated.

Descriptive of the total sample

In the final set (N = 320) respondents aged from min = 17 to max = 67 took part in the survey. Participants were relatively young since there was a mean age = 27. This could be explained by the fact that the survey was spread on social media, among which a Facebook page of UvA Business Administration students. Among the respondents there were 181 men (56.6%) and 139 (43.4%) women, the biggest share of respondents were Dutch (34.4 %) following by respondents from the United Kingdom (25.0%) and the United States (7.2%). The different nationalities of the respondents can be seen in table 2.

Table 2.

Various nationalities of respondents.

Descriptive of different groups

In this study respondents were randomly assigned to one of the four conditions, differing in price and quality level of the brand extension. The four conditions are tested on equality among different background variables and other factors that could jeopardize the equality of the conditions. There was a fairly equal division of sample size, gender and age among the four conditions. An overview of the mean age, division of and gender and nationalities per condition can be found in table 3. In condition 1 and 4 there was a higher percentage of men compared to conditions 2 and 3. However, these differences in percentage are not significant. The different nationalities represented were fairly equal in the different conditions.

Nationality Frequency Percent Dutch 110 34.4 USA 23 7.2 Australian 13 4.1 UK 80 25.0 Germany 10 3.1 Irish 9 2.8 Italian 6 1.9 Polish 9 2.8 Swedish 5 1.6 European else 26 8.1 Non-European else 29 9.1 Total 320 100.0

Table 3.

Sample descriptive of different levels of price/quality increase of the step-up brand extension.

The different conditions were also tested on equality of amount of current brand owners and non-owners and on number of respondents who were brand owners or non-owners in the past. The different levels of step-up brand extensions all have more non-owners than brand owners. As can be seen in table 4. However, the differences between these groups in owners and non-owners do not differ significantly from each other. Also, the different levels of step-up brand extensions all have more past brand owners than people that didn’t own a puma brand in the past. Nevertheless, once again, the differences between these groups in owners and non-owners do not differ significantly from each other. Since the distribution between the conditions is equal, the unequal division of brand ownership are expected to have no effect on the results of this research.

Level of price/quality increase Males Females Nationality Mean age Low (+25%) 49 33 29 Dutch 19 UK 7 USA 27 rest Total 82 25.88 Moderate (+50%) 43 31 20 Dutch 19 UK 9 USA 26 rest Total 74 27.20 High (+200%) 40 39 27 Dutch 23 UK 4 USA 25 rest Total 79 27.34

Very high (+500%) 49 36 34 Dutch

19 UK 3 USA 29 rest

Table 4.

Sample descriptives of the current and previous brand ownership and non-ownership distribution over the four different conditions

Level of price/quality increase

Ownership Ownership past Total N

No Yes No Yes 1. Low 64 15 31 48 79 2. Moderate 64 10 29 45 74 3. High 66 16 36 46 82 4. Very high 71 14 30 55 85 Total 265 55 126 194 320

Furthermore, the different levels of step-up brand extensions and accompanying levels of affective commitments with the Puma brand within the groups were measured, an overview of these measures can be seen in table 5. The differences between these groups in affective commitment did not differ significantly from each other.

Table 5.

Sample descriptives of the level of affective commitment toward the Puma brand and its distribution over the four different conditions

Level of price/quality increase N Mean St. Dev 1. Low 79 3.42 1.17 2. Moderate 74 3.50 1.18 3. High 82 3.39 1.18 4. Very high 85 3.38 .91 Total 320 3.42 1.11

Lastly, before the manipulation in the experiment, the perception of the market position of the Puma brand was checked on equality between the different conditions. The perceived levels of market position were for every condition around 4, which indicates that there is no sign of a ceiling effect for this brand. Furthermore the differences between these groups in perceived market position did not significantly differ from each other. An overview of these results can be found in table 6.

Table 6.

Sample descriptives of the level of perceived market position of the parent brand Puma and its distribution over the four different conditions

Level of price/quality increase N Mean St. Dev 1. Low 79 4.18 1.03 2. Moderate 74 4.07 1.07 3. High 82 3.97 .93 4. Very high 85 4.08 .98 Total 320 4.08 1.00

Correlations

A correlation matrix (table 7) was created in order to test for correlations among the variables and check whether these correlations were as expected or alarming. Moreover, average scores and standard deviations per variable were inspected. The variables: parent brand evaluation, purchase intentions and perceived value correlated positively. According to theory these constructs are related. Noteworthy is that familiarity and ownership are also positively correlated to parent brand evaluation (and purchase intentions and perceived value), which could indicate that experience with the brand is related to a more positive evaluation. Interestingly, the ownership of Puma in the past has no effect on affective commitment towards the parent brand. A possible explanation for this phenomenon is consumers having negative past experiences with the brand. Lastly, the quality evaluation of the brand extension and the increase in the four conditions was significantly positively correlated, which would be in line with the expectations. Since the four conditions contain an increase in quality of the brand extension.

Table 7.

Pearson correlation coefficients, means and standard deviations for different variables

Note: Correlation significant with p ≤ .05; **Correlation is significant with p ≤ .01; N = 320

Manipulations test

In the four different conditions of this study, there was a manipulation of the quality and price of a fictitious newly introduced brand extension. The four conditions had accordingly an increase in price of 25%, 50%, 200% and 500%. The manipulation was followed up by a control question, questioning the perceived quality of the brand extension. In order to test for the difference in quality in the conditions through levels of the control question, a one-way ANOVA was computed. For this manipulation check, the perceived level of quality of the brand extension was the dependent variable. A one-way ANOVA could be computed since the variable of perceived quality is continuous. Levene’s test of homoscedasticity showed that the variances for the quality of the brand extension were equal, F(3, 316) = 2.59, p = .335.

The analysis indicated that there was no statistically significant effect at a 5% level of the perceived quality on the levels of manipulation of the brand extensions, F(3, 316) = 2.89, p = .053. However, results of this test are close to the threshold of significance. Tukey post-hoc tests revealed that the perceived quality of the extension was significantly higher in the level 4 group compared to the level 1 group (p < .05). There was no statistically significant difference of level 1 with level 2 (p =

M SD 1 2 3 4 5 6 7 8 9

1. Brand evaluation 4.69 .97 1

2. Evaluation quality of extension 5.33 1.05 .408** 1

3. Familiarity 5.67 1.46 .228** .203** 1

4. Purchase intention 4.33 1.60 .667* .270** .311** 1

5. Perceived value 4.65 1.13 .703** .315** .236** .647** 1

6. Affective Commitment 3.42 1.11 .562** .324** .291** .534** .491** 1

7. Brand ownership .17 .38 .187** 0.001 .206** .288** .236** .292** 1 8. Brand ownership in the past .61 .50 .76** 0.047 .222* .315** .307** -0.024 .250** 1 9. Levels of price/quality of the extension 2.54 1.13 -0.015 .147** 0.063 0.009 -0.010 -.01 -0.005 -.05 1

.37), level 1 with level 3 (p = .22), level 2 with level 3 (p = .98), level 2 with level 4 (p = .74) and level 3 with level 4 (p = .89). These results indicate that the manipulation of the quality, partially didn’t succeed. Technically future results can only be based on condition 1 and 4. However, the manipulation of price obviously still holds, therefore there still is a difference between these conditions based on price.

The effect of gradual increase in step-up brand extensions on parent brand

evaluation

A one-way ANOVA was computed to test the differences in parent brand evaluation after the introduction of a step-up brand extension, differing in quality and price. The one-way ANOVA is able to show the differences in the mean performance of the different conditions on the dependent variable, parent brand evaluation.

There was no statistically significant effect of the brand extension levels on the total brand evaluation of the parent brand, F(3, 316)=.03, p = .99. Point of interest for this study however is the differences in parent brand evaluations per conditions, to examine whether mean levels of parent brand evaluations are higher for a small increase in price and quality of the brand extension and whether mean scores get lower again for the highest levels in price and quality of the extension. In table 8 can be seen what the different average scores on total brand evaluations are. In the table can be seen that condition 1 scores highest on parent brand evaluations (M = 4.71, SD = 1.08), and condition 4 scores lowest (M = 4.66, SD = .73). Even though these scores do not significantly differ, it could be an indicator of a more positive parent brand evaluation for a smaller step-up brand extension that for a big step-up brand extension. It cannot be concluded where the turning point in evaluation is, since condition 3 evaluated the parent brand more favorably (M = 4.70, SD = .96) than condition 2 (M = 4.68, SD = 1.11).

Table 8.

Different total parent brand evaluations between the four levels of step-up brand extensions Level of price/quality increase N Mean St. Dev 1. Low 79 4.71 1.08 2. Moderate 74 4.68 1.11 3. High 82 4.70 .96 4. Very high 85 4.66 .73 Total 320 4.67 0.97

Additional analysis: the effect of perceived quality of the step-up brand

extension on parent brand evaluation

In abovementioned confirmative analysis no significant impact (negative nor positive) of different levels of step-up brand extension on the evaluation of the parent brand was found, therefore an additional analysis was computed. In this analysis the evaluation respondents gave of the step-up brand extensions (as a control question) is taken as a starter point to test the effect of the differences in perceived quality of a step-up extension on the parent brand. A linear regression analysis was performed to investigate the effect that a perceived quality of a step-up brand extension has on the evaluation of the parent brand.

One predictor of parent brand evaluation was entered: perceived quality of the brand extension, after controlling for age and gender. In the first step of this hierarchical multiple regression, two predictors were entered: age and gender. This model was statistically significant F(2, 317) = 7.82, p < .01 and explains 4.1% of the variance in parent brand evaluation. After entry of perceived quality of the brand extension at step 2 the total variance explained by the model as a whole was 19.4% F(1, 316) = 61.03, p < .01. The introduction of perceived quality of the brand extension explained additional 15.4% variance in parent brand evaluation, after controlling for gender and age (R² change = .15; F(2, 209) = 61.03, p < .01). Perceived quality of the extension was statistically significant with a higher Beta value (Beta = .39, p < .01) than gender (Beta = .15, p < .05) and age (Beta = .12, p < .05). In other words, if perceived quality of a brand extension increases for one, the parent brand evaluation increases with 0.39. An overview of these results can be viewed in table 9.

Table 9.

Hierarchical regression model of perceived quality of step-up brand extension on parent brand evaluation, controlled for age and gender.

R R² R² change B SE Beta t Step 1 .22 .05*** Age .02 .01 .14* 2.46 Gender .34 .12 .17** 3.11 Step 2 .45 .20*** .15*** Age .02 .01 .12* 2.31 Gender .29 .10 .15** 2.92 Perceived quality brand extension .36 .05 .39*** 7.81

Note. Statistical significance: * p < .05; ** p < .01; *** p < .001

To interpret the relationship between perceived quality of the step-up brand extension and the parent brand evaluation, the nature of this relationship has to be investigated. In figure 2 the representation of the relationship between step-up brand extension and the parent brand evaluation is visible. The trend line shows a linear relationship between the two variables. This means that with the gradual increase of quality perceptions of the step-up brand extension the evaluation of the parent brand increases. This linear relationship shows no indication of a turning point into a less positive (or even negative) effect of the increase in perceived quality on parent brand evaluation. Thus, contradictory to what was expected in H1, no negative evaluations of the parent brand occur after the introduction of a step-up brand extension. This analysis supports on top of the previous confirmative analysis to reject H1.

Figure 2. Linear relationship between perceived quality of the step-up brand extension and the evaluation of the parent brand.

The effect of brand ownership on the relationship of gradual increase in step-up brand extensions on parent brand evaluation

A factorial ANOVA was computed to test the main effect: difference in parent brand evaluation after the introduction of a brand extension, differing in quality and price. And to test the interaction effect: to indicate if being an owner of Puma footwear has an effect on this relationship. This test was conducted since there were two categorical independent variables (levels of price/quality and brand ownership) and there was one numerical dependent variable (parent brand evaluation). The factorial univariate ANOVA is able to test underlying relationships between the variables. Results can be found in table 10.

Table 10.

Results of factorial ANOVA, the effect of different price and quality levels of a brand extension on the evaluation of the parent brand, and the moderating effect of brand ownership.

Source Mean Square F Sig. Partial Eta

Squared Levels of brand extensions .891 .973 .405 .009 Ownership 10.632 11.621 .001 .036 Levels * Ownership 1.583 1.730 .161 .016

There was no significant main effect of levels of quality/price on evaluation of the parent brand, F(3, 312) = .97, p = 0.41, η2 = .01. Despite the insignificant effect a Tukey post-hoc test was conducted to gain extra information about the specific differences between the conditions. The Tukey post-hoc tests revealed that there were no significant differences on parent brand evaluation for the different price/quality groups. In other words, the introduction of a step-up brand extension has no effect on the evaluation the parent brand. Regardless of the distance between the extension and the parent brand product. Therefore, H1 was not supported.

The effect of brand ownership on parent brand evaluation was also measured in a factorial ANOVA. There was a significant main effect of brand ownership on parent brand evaluation, F(1, 312) = 11.62, p < .05, η2 = .04. Ownership of the Puma brand leads to higher evaluations as opposed to non-ownership.

There was a non-significant interaction effect between the levels of price/quality of the brand extension and brand ownership F(3, 312) = 1.73, p = .16, η2 = .02. In this factorial ANOVA I did not control for age, sex and nationality since there were no significant differences among these factors between the different conditions.

Another factorial ANOVA was computed to test the effect of past ownership of the Puma brand on the relationship between the different levels of step-up brand extensions on the parent brand evaluation. There was a significant main effect of past brand ownership on parent brand evaluation, F(1,311) = 10.47, p < .05, η2 = .03. Again, parent brand evaluations were higher for past Puma owners than for non-owners.

There was a non-significant interaction effect between the levels of price/quality of the brand extension and past brand ownership, F(3,311) = 1.36, p = .26, η2 = .01. An overview of these results can be found in table 11.

Table 11.

Results of factorial ANOVA, the effect of different price and quality levels of a brand extension on the evaluation of the parent brand, and the moderating effect of past brand ownership.

Source Mean

Square

F Sig. Partial Eta

Squared Levels of brand extensions .017 .019 .996 .000 Past ownership 9.623 10.465 .001 .033 Levels * Past ownership 1.251 1.361 .255 .013

In sum, from the analysis was found that the introduction of a step-up parent brand extension had no effect on the evaluation of the parent brand, therefore hypothesis H1 was not supported. Brand ownership has a direct significant effect on parent brand evaluation, which supports that experience with the Puma brand has a positive effect on the evaluation of the parent brand. However, (past) brand ownership did not influence the effect that step-up brand extensions have on parent brand evaluation, this means that even brand ownership does not bring forward the expected effect that step-up brand extension are expected to have on the parent brand evaluation. Therefore, neither hypothesis H2 was not supported.

Additional analyses

Despite expectations based on previous studies, there are no significant effects found in the confirmative analyses of the different price and quality levels of step up brand extensions on parent brand evaluation, subsequently no moderated effects were found for brand ownership. Accordingly, additional analyses were conducted to find effects for different variables that I measured additionally in the experiment.

Purchase intentions and perceived value of the parent brand

A factorial ANOVA was conducted to bring forward the effect of the levels of a step-up brand extension on the aspect of purchase intentions of the parent brand. No significant effect was found, F(3, 316) = .15, p = .93. Furthermore, a factorial ANOVA was computed to reveal the effect of the

levels of step-up brand extensions on the perceived value of the parent brand, no significant effect for this relation was found, F(3,314) = .07 , p = .97. These analyses were done with dependent variables that are closely related to the variable parent brand evaluations. These results confirm that the absence of effect on parent brand evaluation are not due to errors in the measurement of parent brand evaluation.

Affective commitment

Another factorial ANOVA was computed to test the effect of affective commitment towards the Puma brand on the relationship between the different levels of step-up brand extensions on the parent brand evaluation. Affective commitment towards the parent brand was included in the experiment as a more elaborate alternative to brand ownership. In order to test this relationship, affective commitment was divided into two categories based on the 7 point Likert answer scale: low affective commitment (scoring 1-3.99) and high affective commitment (scoring 4.01 – 7). Respondents that scored exactly average (4.0) were excluded from the analysis. The categories were divided as displayed in table 12.

Table 12.

Number of respondents per category of affective commitment.

Level of affective commitment N Percentage

1. Low 214 73.8

2. High 76 26.2

Total 290

Missing 30

Total 320

In line with previous factorial ANOVA’s, there was no main effect of the levels of step-up brand extension on the parent brand evaluation.

However, there was a significant main effect of affective brand commitment on parent brand evaluation, F(1, 282) = 61.31, p < .05, η2 = .18. A higher level of affective commitment is positively related to an increase in parent brand evaluation. There was also a significant interaction effect found between the levels of price/quality of the brand extension and affective commitment on parent brand

evaluation, F(3, 282) = 2.18, p < .05, η2 = .03. The factorial ANOVA revealed the relationship between levels of step-up brand extensions and parent brand evaluation is the dependent on the level of affective commitment.

To reveal the differences of different parent brand evaluations per level of step-up brand extension within the two categories of affective commitment, a pairwise comparison was computed. Per category of affective commitment, a one-way ANOVA was conducted. The Tukey’s post-hoc tests per category of affective commitment showed the different scores per levels of step-up brand extensions. The results show no statistically significant results between the levels of step-up brand extensions. The results of these tests can be seen in table 13.

The insignificance of the differences between the levels of step-up brand extensions per category of affective commitment show that the parent brand evaluations for consumers with low affective commitment are not different for the various levels of step-up brand extensions. A similar effect appears for consumers who have a high affective commitment towards the brand.

Table 13.

Results of the post-hoc tests for a factorial ANOVA, the effect of different price and quality levels of a brand extension on the evaluation of the parent brand, and the moderating effect of two categories of affective commitment

Level of affective commitment

Levels of step-up

brand extension Mean

Std.

Deviation N Low affective

commitment 1. Low _{2. Moderate} 4.26 _4.46 0.96 _0.96 50 ₄₅

3. High 4.44 0.85 58

4. Very high 4.51 0.76 61

Total 4.42 0.88 214

High affective

commitment 1. Low _{2. Moderate} 5.57 _5.03 0.73 _1.30 23 ₁₈

3. High 5.66 0.81 16

4. Very high 5.10 0.45 19

Total 5.34 0.89 76

Note. All results were non-significant.

There is a direct effect of affective commitment which shows that the higher affective commitment towards the brand, the higher the evaluation of the parent brand. Moreover an interaction effect was

found, which shows that the effect of an introduction of a step-up brand extension on parent brand evaluation is dependent on the level on the value of affective commitment. Results within the categories of low affective commitment indicate no noteworthy differences between the levels of step-up extensions. Within the category of high affective commitment a zigzag shaped relationship between the levels of step-up brand extension on parent brand evaluation is visible. However, the differences between the different step-up brand extensions are non-significant so no conclusion can be drawn.

These relationships were not hypothesized, though affective commitment is closely related to brand ownership. Since there were no differences found in brand evaluation on the different levels of step-up brand extensions, this analysis supports to reject H2.