Examining brand extensions and their potential to dilute team brand associations.

Walsh, Patrick ; Ross, Stephen D.

Introduction

When an organization uses a pre-established brand name to enter a new product class the new product is known as a brand extension (Aaker, 1991). For example, when a professional sports team opens a team merchandise store, they have extended their brand from their primary product category of professional sports and entertainment into the new product category of retail. While this may sound similar to licensed products, there is an important distinction between the two. Licensing occurs when an organization allows a separate company to utilize their brand image, logo likeness, etc. for a fee on their products. For example, Reebok is the official on-field licensed apparel company of the NFL. As such, they can utilize the NFL brand assets in developing on-field merchandise and then sell that merchandise at retail. A brand extension, however, does not involve any outside companies. In this instance the team itself develops the new product or service without outside involvement. In most instances, the team merchandise store example above is a brand extension as the store is typically created by the team.

As costs to operate teams continue to rise, organizations will seek new and innovative ways of increasing revenue for the team outside of the core product, the game itself. Brand extensions are a common revenue generation tactic as they allow a team to enter a new product category while taking advantage of a pre-existing brand name and the image associated with that parent brand name (e.g., the team's brand). In fact, it has been suggested that the introduction and management of brand extensions such as sports grills and merchandise stores will increase as a practice for sport teams (Gladden, Irwin, & Sutton, 2001). In addition to having a positive impact on the team's revenue, brand extensions may enhance the brand image of the team and also provide an avenue for the fans to interact with the team's brand outside of just attending or watching one of the team's events.

While brand extensions have the potential to result in increased revenue and brand interaction for the team, they also have the potential to fail and harm the brand equity that has been developed by the team (John, Loken, & Joiner, 1998; Loken & John, 1993). In addition, there is great financial risk when introducing brand extensions. In February of 2006 ESPN introduced a new mobile phone service (Mobile ESPN) which provided cell phone service with sports content to its subscribers. An estimated $150 million was spent to introduce this brand extension, which debuted with advertisements during Super Bowl XL (Fisher & Ourand, 2006). In the end, consumers were not responsive and subscriptions ranged from an estimated 5,000 to 20,000 subscribers, well short of ESPN's estimate of 250,000 subscriptions (Fisher & Ourand, 2006). As a result, ESPN discontinued Mobile ESPN in December of 2006, less than one year after the service debuted.

The Mobile ESPN example illustrates just how risky a failed brand extension can be financially, which could ultimately impact the equity of the brand that introduces the extension. Any damage in team brand equity could impact the team's ability to generate revenue, and potentially decrease a fan's overall behavior and attitudes toward the team (Fink, Trail, & Anderson, 2002; Funk & James, 2001; Mahony, Madrigal, & Howard, 2000; Ross, 2006; Trail, Fink, & Anderson, 2000; Trail & James, 2001). As such, it is important that sport marketers understand both the positive and negative consequences associated with developing and offering brand extensions. However, despite the growing number of brand extensions in professional sport, very little empirical research has been conducted on this topic. Most research on brand extensions focuses on the consumer's evaluation of the extension as opposed to the effects that the extension could have on the parent brand.

Consumer Evaluations of Brand Extensions

How consumers evaluate brand extensions relies heavily on their evaluation of the parent brand that is introducing the extensions, and the various components of the parent brand's equity. It is generally supported that a consumer's evaluation of a brand extension is driven by the attitudes and associations one has for the parent brand (Aaker & Keller, 1990; Bhat & Reddy, 2001; Broniarczyk & Alba, 1994). If a consumer has favorable attitudes toward a team and believes the team has a positive brand image then they will transfer these favorable attitudes toward the brand extension, thus resulting in a greater chance to influence purchase. In addition, the perceived quality of the parent brand that introduces the extension will impact a consumer's evaluation of an extension (Bottomley & Doyle, 1996; Bottomley & Holden, 2001). In this instance, if a professional sport team is considered to have a high-quality product on the field, consumers are more likely to evaluate that team's brand extension as being high in quality.

However, that is not to say that this would occur for every brand extension that a team might introduce, as other factors could influence this effect. For example, the fit between the product class of the parent brand and the product class of the brand extension will have an impact on how consumers evaluate a team's brand extension. In general, the greater the perceived fit the more likely it is that the consumer will have a positive evaluation of the extension (Aaker & Keller, 1990; Apostolopoulou, 2002b; Bhat & Reddy, 2001; Bottomley & Doyle, 1996; Bottomley & Holden, 2001; Volckner & Sattler, 2006). For example, it is likely that consumers would view a brand extension of footballs to be a perceived fit for a professional football team. However, if that same team were to introduce a line of cologne it would likely not be viewed as a perceived fit with the team. Research would indicate that consumers would form more favorable attitudes toward the football extension as the product category of the parent brand (the team) as the product category of footballs share more similar associations than cologne.

While there has been little research on brand extensions in sport, the primary focus has also centered on consumer evaluations of the extensions. Apostolopoulou (2002a) was the first to contribute significantly to the study of brand extensions in sport and categorized sport brand extensions into five different categories: sport related, entertainment related, media related, information related, and low perceived fit extensions. Further research on team related extensions determined that loyal fans of a team are more likely to have favorable attitudes and opinions about a team's brand extension as opposed to those at lower levels of identification or loyalty (Apostolopoulou, 2002b). It was also noted that this effect occurs regardless of the perceived fit between the team and the brand extension (Apostolopoulou, 2002b). This is most likely occurring as highly identified fans access more brand associations (Ross & James, 2007), and are more likely to form strong attitudes toward their team, which are resistant to change (Funk & James, 2001; Mahony, Madrigal, & Howard, 2000). Finally, Papdimitriou, Apostolopoulou, and Loukas (2004) found that perceived fit is higher for sport related extensions (e.g., team camps, sporting equipment, etc.) and consumers are more likely to purchase team related extensions, which are a higher perceived fit. While the previous research indicates the parent brand has an effect on the consumer's evaluation of a brand extension, very little research has focused on the actual effects that brand extensions have on parent brands. One way to examine this potential impact on the parent brand would be to determine if there are any changes to team brand associations after exposure to a brand extension.

Team Brand Associations

Brand associations are the thoughts and ideas that an individual has in their memory for a particular good or service (Aaker, 1996; Keller, 1993). Gladden and Funk (2002) were the first to attempt to identify what brand associations one might have for their favorite sports team. They proposed that there were 16 different association dimensions, which could further be classified similar to Keller's (1993) model of brand equity as attribute, benefit, or attitude associations. The attribute associations include Success (win/loss record, making the playoffs, and winning championships), Head Coach (coach is well known/liked, does a good job), Star Player (like to watch team's star players, team

has a star player), Management (competent front office), Stadium (attractive stadium, has character, adds to enjoyment of game), Logo Design (nice logo, colors, and uniforms), Product Delivery (games are exciting, entertaining, and enjoyable), and Tradition (winning, rich team history). Gladden and Funk's (2002) benefit associations include Identification (level of connection with the team), Nostalgia (team's ability to create and bring back memories), Pride in Place (team elevates image of community), Escape (diversion from everyday life), and Peer Group Acceptance (influence of friends in following team). Finally, the attitude associations include the team's importance or personal relevance to the fan, the fan's self-perceived knowledge of the team, and affect.

Ross et al. (2006) built off of Gladden and Funk's (2002) classification of team brand associations and developed the team brand association scale (TBAS). This scale, which has been utilized and tested in a variety of sport settings (Ross, Bang, & Lee, 2007; Ross, 2007), measures 11 different team brand association dimensions: the non-player personnel associated with a particular team (Non-player Personnel); 2) the quality, success, and performance of the team (Team Success); 3) the history surrounding the team (Team History); 4) the stadium and community in which the team plays (Stadium Community); 5) the specific characteristics that team displays during a game (Team Play Characteristics); 6) the identifying marks associated with the team (Brand Mark); 7) the degree to which an organization is committed to its fans and community (Organizational Attributes); 8) consuming the food and beverage at the stadium (Concessions); 9) associating with friends or other fans of the team (Social Interaction); 10) the most significant competitors of the team (Rivalry); and 11) the perceived level of connection or enduring affiliation that the fan base has with a particular team (Commitment).

It has been suggested that brand associations are one of the vital factors that comprise sport spectator-based brand equity, and that associations could have an impact on the spectators' loyalty toward a team, the excitement or level of entertainment that spectators experience at a sporting event, the ability for the team to solicit revenue from fans and sponsors, and the opportunity to develop brand extensions (Ross, 2006). Several notable authors (Aaker, 1996; Berry, 2000; Keller, 1993) have developed brand equity models; each having brand associations as a key component. In view of this, brand associations are considered to be a major component of brand equity, and thus will be examined in the current study using the TBAS.

Theoretical Background and Hypotheses

Categorization indicates that when consumers encounter a new product in the marketplace they will utilize their previous experience with and/or perceived knowledge of the product's category to make summary judgments about the new product (Loken, 2006). For example, Callaway Golf s line of Big Bertha clubs may be associated with being easy to use and increasing a player's driving distance. When a new line of Big Bertha golf clubs is introduced, consumers assume that the new clubs also share these attributes. This process allows consumers to quickly process information about the new product, leading to attitude formation.

It has been suggested that acceptance of a new product would rely on a variety of factors. For example, the greater the perceived fit, the more likely it is for the new product to be accepted by the consumer (Aaker & Keller, 1990; Bhat & Reddy, 2001; Bottomley & Doyle, 1996; Bottomley & Holden, 2001; Boush & Loken, 1991; Volckner & Sattler, 2006). Category inferences are more likely to be drawn for a new product, which has greater perceived similarity, or fit, between the new product and the parent brand, as they likely hold similar attributes or similar associations (Aaker & Keller, 1990; Bhat & Reddy, 2001; Broniarczyk & Alba, 1994).

In a similar manner, new category members could also have an effect on the beliefs that consumers hold for the existing category. Prior to a study conducted by Loken and John (1993), most research on brand extensions focused on how consumers evaluate brand extensions. Loken and John (1993) designed a study to understand under what circumstances brand extensions are more or less likely to dilute the attribute beliefs that a consumer holds toward the parent brand. Dilution in this instance refers to any damage caused to the brand name by creating undesirable brand associations, damaging perceived quality, or altering existing positive brand associations (Aaker, 1991). Two models from this research that attempt to understand product categories provide the theoretical background for this study, and form the basis for the developed hypotheses focusing on dilution by altering existing positive brand associations.

Bookkeeping Model

The bookkeeping model suggests that a consumer's beliefs about a product will change when they are presented with new inconsistent information (Loken & John, 1993; Weber & Crocker, 1983). This model, similar to many information processing models (Engel & Blackwell, 1992; Petty & Cacioppo, 1986) would then suggest that when a brand extension is introduced to the marketplace, consumers will update their beliefs and attitudes about the parent brand. Therefore, if the new brand extension portrays associations which are not consistent with the parent brand's associations, consumers will modify their beliefs about the parent brand.

A fairly extensive foundation of literature exists that establishes the role of commitment in the information processing capabilities of consumers when exposed to counterattitudinal brand information (Raju, Unnava, & Montgomery, 2009). When applied to team brand extensions, the bookkeeping model would indicate that if a team introduces a brand extension having inconsistent associations, when compared with the team's brand associations, then dilution of the team's brand associations would occur. This dilution is more likely to occur if attitude measurement takes place immediately following exposure to the extension, as the associations for the team are salient in the mind of the consumer, which then encourages piecemeal processing rather than category based processing (Loken & John, 1993). In other words, the consumer is not using their previous knowledge of the category to make judgments about the parent brand, but is using the new information which has just been presented to them when assessing the parent brand's associations. Thus, the bookkeeping model forms the basis for the following hypothesis:

H1: Dilution of team associations will occur when exposed to a brand extension with inconsistent associations regardless of typicality.

Typicality Model

The typicality model proposes that dilution of parent brand's attributes will be more likely to occur when brand extensions are a perceived fit with the parent brand's product category, while at the same time having inconsistent associations. (Loken & John, 1993; Rothbart & Lewis, 1988). For example, if a specific team were to open a merchandise retail store, consumers will be more likely to view this as being typical, as merchandise stores are now in widespread use by sport organizations. However, if the retail store exhibits attributes that are inconsistent (e.g., "blue collar" team selling designer clothing), it is possible for the attributes of the parent brand/teams to suffer from dilution. From this information the following hypothesis was formulated:

H2a: Dilution of team associations will occur when exposed to a brand extension that is a perceived fit yet has incongruent associations.

Conversely, the typicality model also states that a brand extension which is not a perceived fit, and also has inconsistent associations, will not dilute team brand equity (Loken & John, 1993; Rothbart & Lewis, 1988). In this example, if a team with a "blue collar" personality launches a new line of beauty products, it would be very likely that consumers perceive this extension as being very atypical. Thus, according to the typicality model, the atypical extension would have numerous inconsistent attributes, and therefore not dilute the associations for that team. This information discounting effect occurs as the typicality of the brand extension is often the most salient in the mind of the consumer. Loken and John (1993) note that perception of an extremely atypical extension likely induces category-based processing, as opposed to the bookkeeping model where piecemeal processing is often utilized. That is, consumers utilize current knowledge or previous experience with parent brand category to form opinions about the parent brand, as opposed to the new attribute information encountered.

Loken and John (1993) also suggest that the typicality model takes into account the gradient structure of product categories. More specifically, the authors suggest that more typical members of a product category will share more attributes than less typical members. Accordingly, dilution of parent brand attributes is more likely to occur for a brand extension that is considered to be a perceived fit within a given category, but has inconsistent attribute information. However, an extension with inconsistent attributes and perceived not to fit will be less likely to dilute parent brand attributes than the extension which is a perceived fit with inconsistent attributes. Based on the preceding information the following hypothesis was formulated:

H2b: Dilution of team associations will not occur when exposed to a brand extension that is not a perceived fit and has incongruent associations.

Proposed Covariate

Given that psychological level of commitment toward a team has been shown to influence a number of behavior outcomes, team identification is proposed here to be a covariate. Team identification has been shown to have a strong positive relationship to sport fan consumptive behavior (Fink et al., 2002), while Wann and Branscombe (1993) determined that highly identified fans are more likely to attend games and spend time and money to watch their favorite team play. In addition, research has suggested a significant relationship between identification and the motives for attending sporting events (Trail et al., 2000; Trail & James, 2001). Furthermore, highly identified fans are suggested to be less likely to change attitudes and behaviors toward a team as compared to those with lower levels of identification (Funk & James, 2001; Mahony et al., 2000). That is, highly identified fans often form a strong relationship with their team, to the point where the team has become an integral part of their lives, and thus strongly influences attitudes and behaviors.

In terms of brand extensions, Apostolopoulou (2002b) found that team identification had a significant positive effect on a fan's evaluation of a sport related brand extension regardless of typicality. Based on these findings, it is possible that a similar relationship might exist when fans with varying levels of identification evaluate beliefs about the team's brand. In a related study, Ross and James (2007) found that highly identified fans access a greater number of brand associations regarding their favorite team than those individuals at lower levels of identification. It could, therefore, be suggested that those with higher identification levels might be more motivated to process the information presented to them.

Motivation to process information has been shown to moderate the effect that brand extensions have on the evaluation of a parent brand (Gurhan-Canli & Maheswaran, 1998). When exposed to hypothetical brand extensions with varying associations, highly identified fans will likely access more brand associations than moderately and low identified fans. Highly identified fans will, therefore, be more motivated to process extension information, resulting in a greater number of associations being activated within the consumer's neural network of associations.

Related literature on the similar constructs of brand ownership and brand commitment also support the notion of team identification as a proposed covariate. Kirmani, Sood, and Bridges (1999) determined that individuals who own a particular brand tend to have more familiarity, knowledge, and involvement with the brand. In addition, owners of brands tend to have more favorable responses to extensions introduced from the owned brand (Kirmani et al., 1999). Research has also suggested that highly committed consumers will discount negative information about a brand as compared to less committed consumers, thereby reducing the likelihood of attitude dilution (Ahlulwalia, Burnkrant, & Unnava, 2000; Ahluwalia, Unnava, & Burnkrant, 2001). Based upon this review of literature the following hypothesis was formulated:

H3: Dilution of team associations will be less likely to occur for highly identified fans than fans at moderate or low levels of identification, and less likely to occur for moderately identified fans than fans at low levels of identification, regardless of typicality and congruency of associations.

Research Method

To examine how the introduction of brand extensions of a professional sports team impacts the brand associations of the parent brand/team two studies were conducted. Study One was designed with the purpose of obtaining the necessary information to conduct an experimental study in Study Two. Study Two was a post-test only control experimental design as participants were randomly assigned to groups, a treatment was delivered to the experimental groups, and then the control group measures were compared to the experimental groups (Creswell, 2003). The population of interest for the current study was spectators of a National Hockey League (NHL) team based in the Midwest region of the United States.

Study One

Utilizing a random sampling procedure, 1,500 individuals were selected from the team's ticket database to participate in Study One, which consisted of a self-administered web-based survey. The participants (n=266) were asked to respond to a variety of items to obtain the necessary information to conduct Study Two. First, participants responded to the 41-item TBAS (Ross et al., 2006) by indicating their agreement on a scale of 1 (Disagree) to 7 (Agree) that the various brand association dimension items were characteristics of the team. Similar to the study conducted by Loken and John (1993), the respondents also responded to a scale of 1 (Dissimilar) to 7 (Similar) on how similar (e.g., perceived fit) the image of 15 different product categories was with the image of the team. The 15 categories were similar to the extensions utilized by Papadimitriou and Apostolopoulou (2004), and provide a wide range of items which may or may not be typically introduced by professional sports teams. The items included a bar and grill, mobile phone service, youth hockey camp, cosmetics, sport energy drink, candy bar, hockey equipment, travel magazine, sports magazine, designer clothing, bottled water, sports news television show, energy bar, pizza delivery, and exercise equipment. Finally, four questions were included to measure level of team identification of the pre-test respondents. The four items examined team identification through 1) the feeling of ownership of a particular team, 2) thinking of oneself as a loyal fan of that team, 3) the importance one puts on being a fan of the team, and 4) the importance of expressing to others that they are a fan of the team (James & Ross, 2002).

The results of Study One allowed for the understanding of the most salient associations that fans held for the team. Specifically, those attributes from the TBAS which had the highest mean scores were utilized as the congruent attributes, team success (M=5.78) and commitment (M=6.25), while those with the lowest mean scores were deemed as being the most incongruent attributes, team history (M=4.96) and social interaction (M=4.88). While the team history and social interaction dimensions were above the midpoint of the scale, they were the associations rated the lowest among all examined dimensions and considered to be the most incongruent in the context of the current study. In addition, the results from Study One helped determine which two product categories would be utilized as the hypothetical team brand extensions in the experimental treatment. One product with a high mean score was utilized as the typical/high perceived fit extension, Hockey Equipment (M=6.27), while another product with a low mean score was identified as the atypical/low perceived fit brand extension, Designer Clothing (M=2.39). Finally, it was found that the majority of the individuals in the team's database were highly identified fans, and would therefore not exhibit enough variation to examine the covariation effect of team identification. As such, a supplementary sample was required in Study Two to fully analyze the impact that identification was having on brand extension evaluation.

Study Two

Following data collection and analysis of Study One, a new random sample of 700 individuals was selected from the team's database. These 700 potential participants were then randomly assigned into either the control group or one of the four experimental groups. In addition, as discovered in Study One it was determined that the majority of individuals in the team's database would be classified as highly identified fans. Research has indicated that highly identified fans are less likely to moderate their attitudes toward their favorite teams (Funk & James, 2001; Mahony et al., 2000). Therefore, in order to gain a more representative sample of a fan population, and to properly address the proposed hypotheses, a convenience sample of 316 students from a nearby university were solicited as participants in the study. This particular group of students attended a university in the same market as the NHL team utilized in the study. It is likely that some of the students were fans of the team, and expected that their levels of identification will be less than those who exist in the team's list given the database consisted of fans who have purchased tickets in the past and have opted to receive information from the team. In addition, the use of students was determined to be appropriate as they mirror the teams target market and are "significant consumers and users of sport" (Ross, 2006, p. 265).

The control group responded to the TBAS without any exposure to a brand extension, and thus provided a baseline measure of the participant's team brand associations. Based on the results of the survey from study one, two extension product categories (Hockey Equipment and Designer Clothing) served as the basis for the experimental group treatments and were manipulated to have either congruent attributes or incongruent attributes. The four experimental groups therefore consisted of: 1) Typical brand extension with congruent associations, 2) Typical brand extension with incongruent associations, 3) Atypical brand extension with congruent associations, and 4) Atypical brand extension with incongruent associations. Before completing the survey, each participant was exposed to one of the four hypothetical brand extensions by being asked to read a statement that utilized the appropriate associations in the wording. The statement represented a marketing message that was similar to what a team might utilize, and fans would be subsequently exposed to, when introducing a new product to the market. The statements described the brand extension by name, indicated that the team was planning on introducing the extension, and then how potential users could benefit from the new extension.

Multivariate analysis of variance (MANOVA) was utilized to determine if there were differences in the TBAS dimensions based on the manipulation of the independent variable (brand extensions with varying degrees of typicality and association congruency) when compared to the control group. Dunnett's post hoc analysis was conducted for each of the brand association dimensions comparing the contrasts between the control group and each of the four experimental groups. Multivariate Analysis of Covariance (MANCOVA) was then utilized to determine any effect that team identification may have had on the brand association dimensions. Multiple regression analysis was then conducted to examine the percent of variance in brand association dimensions attributed to respondent level of identification. Finally, where instances of dilution occurred, a univariate analysis of variance (ANOVA) was conducted to examine if dilution was more or less likely to occur for highly identified fans as compared to fans in the moderate and low level identification segments.

Results

Reliability and Validity of the TBAS

A confirmatory factor analysis (CFA) using LISREL 8.54 was conducted to test the validity of the 11 TBAS dimensions in this experimental setting. As suggested by Kline (1998), multiple fit indices were examined. The RMSEA (.073) (90% CI = .068, .078), ECVI (8.07) (90% CI = 7.59, 8.58), TLI (.96), and CFI (.96) all reached either an acceptable or good fit to the data, while GFI (.75) was a marginal fit. In order to assess the internal reliability of the TBAS, Cronbach's alpha coefficients were calculated and ranged from .68 to .89, with just one factor (Rivalry, a=.68) failing to meet Nunally and Bernstein's (1994) acceptable level of .70. The results supported the overall and reliability of the TBAS and ensured that further analysis could be conducted with this scale.

Respondent Profile

Over half (56.5%) of the respondents were males, and between the ages of 18-29 (51.9%). Nearly 40% were between the ages of 30 and 49, with 22.7% between the ages of 30 and 39, and 17.3% between the ages of 40 and 49. The remainder of the respondents were between the ages of 50 and 59 (7.3%), or the ages of 60 and 69 (0.8%). The ethnic profile of the respondents was dominated by white/caucasians who made up 95.7% of the respondents. In addition, the sample was highly educated as 28.6% of the respondents indicated that the highest level of education they have completed was a bachelor's degree, while 10% earned an advanced degree

Analysis of Hypotheses

Using Wilk's criterion, the MANOVA suggested a significant difference between the groups on the team brand association dimensions (A = .617, F (4, 185) = 2.050, p<.01). The brand association ratings across the groups are presented in Table 1. While only the infrequent and frequent associations were utilized in the experimental group scenarios, each of the brand association dimensions was examined for potential dilution effects. Post hoc tests revealed that there was a significant difference in the Social Interaction association dimension between the control group and experimental group four (p < .05). Specifically, respondents in the control group had a significantly higher mean score on the Social Interaction dimension (M = 5.21) than the experimental group (M = 4.31). This indicated that dilution of the Social Interaction dimension did occur, and thus provided partial support for Hypothesis 1, suggesting potential dilution when consumers are exposed to a brand extension with inconsistent attributes. Conversely, some dilution did occur and thus would not support Hypothesis 2b, which suggested that dilution would not occur when exposed to an atypical brand extension with incongruent attributes. Finally, no other significant differences were found, and thus Hypothesis 2a was also not supported.

The MANCOVA results revealed that level of team identification had a significant impact on the brand association ratings across the groups (A = .525, F (4, 184) = 14.316, p<.01). Furthermore, the regression results indicated that across all 11 TBAS dimensions group membership was not significant (p > .10) in predicting the brand association scores while controlling for team identification. However, in each case the team identification segment explained a significant amount of the variance in brand association dimension rating (p < .01). Specifically, the results suggested that the majority of the variance in TBAS scores was attributed to the respondent's level of identification rather than exposure to the hypothetical brand extension.

Finally, in the one instance where dilution did occur, the ANOVA results indicated significant differences in the diluted Social Interaction dimension ratings between the various team identification segments (F(2, 30) = 6.997, p < .01). In particular, the ratings increased between those with low levels of identification (M = 3.46), moderate levels of identification (M = 4.79), and high levels of identification (M = 4.91). Bonferroni post hoc tests revealed that there were significant differences in the ratings between those with high levels of identification and those with low levels of identification (p < .05). In addition, significant differences were found between those with moderate levels of identification and individuals with low levels of identification (p < .05), thus supporting Hypothesis 3.

Discussion and Conclusion

This study represents one of the first efforts to investigate the impact brand extensions have upon the associations of a professional sport team. The results suggest that minimal dilution will occur when introducing brand extensions, and revealed that level of team identification had a significantly greater impact on the team brand than did exposure to the extension. The findings from this research offer a number of important theoretical and managerial implications, as well as providing several avenues for future research.

In summary, effects of dilution did not support the typicality model and showed minimal support for the bookkeeping model. The only evidence of dilution was found when participants were exposed to a non-perceived fit extension with inconsistent associations, thereby contradicting the typicality model. Interestingly, these findings provide some support for the bookkeeping model, suggesting that regardless of typicality, dilution will occur when an extension has inconsistent attribute information (Loken & John, 1993; Weber & Crocker, 1998). It should be noted, however, that both models were developed for the examination of consumer based goods, and this study applied to a service related product of a professional sport team. As partial support was established for the models in the current research, a need exists to develop a deeper understanding of how team related extensions either enhance or dilute the brand associations of professional sport teams.

The results of this study also provide additional empirical support for various frameworks on team identification, attitudinal loyalty, and brand commitment (Ahluwalia et al., 2000; Ahluwalia et al., 2001; Funk & James, 2001; Kirmani et al., 1999; Mahony et al., 2000). That is, the results suggest team identification has a greater impact on a team's brand ratings than simple exposure to hypothetical brand extensions. Furthermore, those individuals with higher levels of identification are less likely to exhibit attitude change toward the team, and as such, dilution is less likely to occur for highly identified fans. While only minor evidence of dilution occurred in the current research, any evidence of dilution is important to understand for sport marketers.

Aaker (1991; 1996) argues that brand associations are often at the root of consumption decisions and typically represent the foundations of brand loyalty formation. Any damage to the perceived brand associations for a team would likely have a direct negative impact on overall brand equity. Any diminished brand equity will impact the team's ability to foster fan loyalty, generate media exposure, and solicit revenue in the form of merchandise sales, ticket sales, and sponsorship agreements (Ross, 2006). Given the consequences of damaged brand equity, it is vitally important for sport marketers to be cautious when introducing brand extensions and protect any beneficial pre-existing brand associations.

Perhaps the most significant and impactful finding of this research is the importance of team identification as it relates to the introduction of brand extensions by a professional sport team. In all instances the majority of the variance in team brand association ratings was attributed to the respondent's level of identification and not exposure to hypothetical brand extensions. In the one instance where minor evidence of dilution did occur, the impact was less likely to occur for highly identified fans than fans at low levels of identification, and also less likely for moderately identified fans than fans at low levels of identification. These results would suggest that while some dilution may occur for teams that introduce brand extensions, loyal fans are not likely to project the dilution to the brand associations of the team.

Professional sport teams can therefore introduce brand extensions supported by positive marketing communication messages without the concern of damaging relationships with the organization's most loyal consumers. As these loyal fans are more likely to purchase tickets, spend money on merchandise, and are more likely to watch their team's game (Fink et al., 2002; Funk & James, 2001; Trail et al., 2000; Trail & James, 2001; Wann & Branscombe, 1993), it is imperative for marketers to understand that extensions will not have a momentous impact on associations and consumptive behavior of loyal fans. As such, teams with a fan base consisting of primarily moderately and highly identified fans can utilize brand extensions without the fear of diluting their brand. This provides teams with another valuable contact point with their core fans and a potential additional revenue-generating source.

Conversely, the findings also suggest that brand extensions are more likely to dilute the associations of individuals with low levels of identification. While these consumers do not engage in significant levels of sport consumptive behavior as it relates to the team (Fink et al., 2002; Funk & James, 2001; Trail & James, 2001; Wann & Branscombe, 1993), they are still an important target market for sport organizations. Sport marketers must be aware that dilution may still occur in less identified fans when introducing brand extensions, and therefore make it more difficult to strengthen the identification levels and consumptive behavior for this group of consumers. Typically, sport teams would like to increase the individual's level of identification in hopes of creating more attitudinal and behavioral loyalty. However, the brand associations for less identified fans are diluted in these situations, and as Ross (2006) suggested, this will have a negative impact on the fan's overall loyalty toward the team.

While the results of this research do indicate a lesser chance of dilution for teams with strong brand loyalty when introducing extensions, it is recommended that teams engage in proper market research prior to launching any extension. Sport marketers must have a thorough understanding of the brand strength and the loyalty levels of the team's fan base ahead of extension introduction. By understanding these characteristics, a sport organization will have a better grasp of the potential impact an extension might have upon the parent brand. If market research demonstrates the organization has a strong loyal fan base, the introduction of extensions may not have any impact on the team's overall associations, and thus could represent a strong revenue generation option for the team. However, a fan base comprised of moderately to low identified individuals might suggest that a team reconsider the extension strategy as the potential exists for increased dilution to team brand associations.

Limitations and Future Research

Given that this study was the first to examine the impact of brand extensions on the brand associations for a professional sport team, there are many potential avenues for future research. First, as with most studies, it is difficult to generalize these results to all professional sport teams given it was conducted with only one team in one sport in one market of the United States. Future research should therefore be conducted to replicate this study and determine if the results are different for teams in different leagues, different markets, and with teams of varied degrees of brand strength and team identification levels. Future research in this area may lend support to the empirical findings of the current study, and aid in the development of managerial strategies and relevant frameworks by which marketers can operate.

As Keller and Aaker (1992) suggested, there is greater potential for attitudinal change among those brands that have weaker core brand ratings. As such, it is recommended here that future research pay special attention to obtaining samples from teams that do not have a strong brand in their respective markets. Future studies should also investigate the impact that extensions have on brand associations among those teams that do not have a strong loyal fan base. The current research was conducted with a team whose fans are predominantly highly identified, and thus may have limited some of the analysis given the brand association dimensions that were utilized as incongruent associations. While these specific dimensions were the lowest rated relative to the other associations, their ratings could be classified as moderately congruent. Additional research should be conducted with teams with comparable numbers of highly identified fans and fans at moderate and low levels of identification. This would provide the researchers with association dimensions which are truly incongruent and present marketers a clearer picture of how extensions are impacting the brand associations of different professional sport teams with varying degrees of brand strength and fan support.

An additional area for future research would be to examine how extensions impact the team's brand when considering different antecedents that impact spectator-based brand equity. In the current study, respondents were exposed to a hypothetical extension with a positive brand claim considered to be organizationally controlled. Future research could focus on impact factors outside the control of the organization and how the factors might go about impacting team brand associations. For example, market induced antecedents such as word of mouth and publicity should be examined. Previous research has indicated that exposure to negative conflicting information with a brand extension will increase the potential for dilution to the parent brand (Ahluwalia & Gurhan-Canli, 2000; Gurhan-Canli & Maheswaran, 1998; John et al., 1998; Loken & John, 1993). By investigating fan exposure to poor word of mouth or negative publicity about an extension, a better understanding of the potential dilution effects will become possible for sport marketers. Additionally, research should focus on experience-based antecedents to the brand extension (i.e., actual use/interaction with the extension). Consumer interaction with an actual brand extension may produce different results as opposed to presenting them with hypothetical information.

References

Aaker, D., & Keller, K. (1990). Consumer evaluations of brand extensions. Journal of Marketing, 54, 27-41.

Aaker, D. (1991). Managing brand equity. New York: Free Press.

Aaker, D. (1996). Building strong brands. New York: Free Press.

Ahluwalia, R., Burnkrant, R. E., & Unnava, H. R. (2000). Consumer response to negative publicity: The moderating role of commitment. Journal of Marketing Research, 37(2), 203-214.

Ahluwalia, R., Unnava, H. R., & Burnkrant, R. E. (2001). The moderating role of commitment on the spillover effect of marketing communications. Journal of Marketing Research, 38(4), 458-470.

Ahluwalia, R., & Gurhan-Canli, Z. (2000). The effects of extensions on the family brand name: An accessibility-diagnosticity perspective. Journal of Consumer Research, 27, 371-381.

Apostolopoulou, A. (2002a). Brand extensions by U.S. professional sport teams: Motivations and keys to success. Sport Marketing Quarterly, 11(4), 205-214.

Apostolopoulou, A. (2002b). The effect of brand strength and perceived fit on the success of sport brand extensions [Abstract]. In the 17th Annual North American Society for Sport Management Conference Abstracts, Canmore, Alberta, Canada, 4.

Berry, L. (2000). Cultivating service brand equity. Journal of the Academy of Marketing Science, 28(1), 128-137.

Bhat, S., & Reddy, S. (2001). The impact of parent brand attribute associations and affect on brand extension evaluation. Journal of Business Research, 53, 111-122.

Bottomley, P., & Doyle, J. (1996). The formation of attitudes towards brand extensions: Testing and generalizing Aaker and Keller's model. International Journal of Research in Marketing, 13, 365-377.

Bottomley, P., & Holden, S. (2001). Do we really know how consumers evaluate brand extensions? Empirical generalizations based on secondary analysis of eight studies. Journal of Marketing Research, 38, 494-500.

Boush, D., & Loken, B. (1991). A process-tracing study of brand extension evaluation. Journal of Marketing Research, 28, 16-28.

Broniarczyk, S., & Alba, J. (1994). The importance of the brand in brand extension. Journal of Marketing Research, 31, 214-228.

Creswell, J. W. (2003). Research design: Qualitative, quantitative, and mixed method approaches. Thousand Oaks, CA: Sage Publications.

Engel, J. F., & Blackwell, R. D. (1992) Consumer behavior. New York: College Publishing.

Fink, J. S., Trail, G. T., & Anderson, D. F. (2002). An examination of team identification: Which motives are most salient to its existence? International Sports Journal, 6(4), 195-207.

Fisher, E., & Ourand, J. (2006, October 2-8). Flop hangs an 'L' on ESPN. SportsBusiness Journal, 9(22), pp. 1, 35.

Funk, D. C., & James, J. (2001). The psychological continuum model: A conceptual framework for understanding an individual's psychological connection to sport. Sport Management Review, 4, 119-150.

Gladden, J., Irwin, R., & Sutton, W. (2001). Managing North American professional sport teams in the new millennium: A focus on building brand equity. Journal of Sport Management, 15, 297-317.

Gurhan-Canli, Z., & Maheswaran, D. (1998). The effect of extensions on brand name dilution and enhancement. Journal of Marketing Research, 35, 464-473.

James, J., & Ross, S. (2002). The motives of sport consumers: Comparisons of major and minor league baseball. International Journal of Sport Management, 3(3), 180-198.

John, D. R., Loken, B., & Joiner, C. (1998). The negative impact of extensions: Can flagship products be diluted? Journal of Marketing, 62, 19-32.

Keller, K., & Aaker, D. (1992). The effects of sequential introduction of brand extensions. Journal of Marketing Research, 23, 35-50.

Keller, K. (1993). Conceptualizing, measuring, and managing customer-based brand equity. Journal of Marketing, 57(1), 1-22.

Kirmani, A., Sood, S., & Bridges, S. (1999). The ownership effect in consumer responses to brand line stretches. Journal of Marketing, 63(1), 88-101

Kline, P. (1998). The new psychometrics: Science, psychology and measurement. London: Routledge.

Loken, B. (2006). Consumer psychology: Categorization, inferences, affect, and persuasion. Annual Review of Psychology, 57, 453-485.

Loken, B., & John, D. R. (1993). Diluting brand beliefs: When do brand extensions have a negative impact? Journal of Marketing, 57, 71-84.

Mahony, D., Madrigal, R., & Howard, D. (2000). Using the psychological commitment to team (PCT) scale to segment sport consumers based on loyalty. Sport Marketing Quarterly, 9(1), 15-25.

Nunnally, J., & Bernstein, I. (1995). Psychometric theory. New York: McGraw Hill.

Papadimitriou, D., Apostolopoulou, A., & Loukas, I. (2004). The role of perceived fit in fans' evaluation of sports brand extensions. International Journal of Sports Marketing and Sponsorship, 6(1), 31-48.

Petty, R. E., & Cacioppo, J. T. (1986) Communication and persuasion: Central and routes to attitude change. New York: Springer Verlag.

Raju, S., Unnava, H., Montgomery, N. (2009). The moderating effect of brand commitment on the evaluation of competitive brands. Journal of Advertising, 38(2), 21-35.

Ross, S. D. (2006). A conceptual framework for understanding spectator-based brand equity. Journal of Sport Management, 20, 22-38.

Ross, S. D., & James, J. (2007). The influence of psychological connection on professional sport team brand associations. International Journal of Sport Management, 8(1), 115-129.

Ross, S. D., James, J., & Vargas, P. (2006). Development of a scale to measure team brand associations in professional sport. Journal of Sport Management, 20(2), 260-279.

Ross, S. D. (2007). Segmenting sport fans using brand associations: A cluster analysis. Sport Marketing Quarterly, 16(1), 15-24.

Ross, S. D., Bang, H., & Lee, S. (2007). Assessing brand associations for intercollegiate ice hockey. Sport Marketing Quarterly, 16(2), 106-114.

Rothbart, M., & Lewis, S. (1988). Inferring category attributes from exemplar attributes: Geometric shapes and social categories. Journal of Personality and Social Psychology, 55, 861-872.

Trail, G. T., Fink, J. S., & Anderson, D. F. (2000). A theoretical model of sport spectator consumption behavior. International Journal of Sport Management, 1, 154-180.

Trail, G. T., & James, J. D. (2001). The motivation scale for sport consumption: Assessment of the scale's psychometric properties. Journal of Sport Behavior, 24(1), 108-127.

Volckner, F., & Sattler, H. (2006). Drivers of brand extension success. Journal of Marketing, 70, 18-34.

Wann, D., & Branscombe, N. (1993). Sports fans: Measuring degree of identification with their team. International Journal of Sport Psychology, 24(1), 1-17.

Weber, R., & Crocker, J. (1983). Cognitive processes in the revision of stereotypic beliefs. Journal of Personality and Social Psychology, 45(5), 961-977.

Patrick Walsh, PhD, is an assistant professor in the Department of Kinesiology at Indiana University. His research interests include sport brand management, new media marketing, and corporate sponsorship. Stephen D. Ross, PhD, is an associate professor of sport management at the University of Minnesota. His research interests include sport brand management, sport consumer psychology, and sport marketing as it relates to the youth segment.

Table 1.

Comparison of TBAS Items Across Groups

TBAS Item Total Control EG 1 EG 2 EG 3 EG 4

Commitment 6.20 6.23 5.85 6.32 6.46 6.02
 (.874) (.814) (1.18) (.809) (.601) (.842)

Stadium Community 5.94 6.13 5.88 5.83 6.11 5.69
 (.915) (.666) (.988) (.920) (.839) (.970)

Logo 5.99 5.89 5.84 5.96 6.11 6.14
 (1.09) (1.01) (1.31) (1.16) (1.09) (.877)

Team Success 5.82 5.66 5.69 5.99 6.01 5.67
 (.885) (.925) (.869) (.879) (.873) (842)

Organizational 5.80 5.77 5.61 5.89 5.98 5.67
 Attributes (.962) (.974) (1.08) (.972) (.824) (.972)

Non-player Personnel 5.57 5.70 5.45 5.50 5.70 5.45
 (.963) (.890) (1.01) (1.03) (.942) (.961)

Team Play 5.52 5.39 5.30 5.69 5.75 5.38
 (1.05) (1.12) (1.15) (.981) (.939) (1.04)

Rivalry 5.28 5.31 5.16 5.29 5.45 5.11
 (.885) (.877) (.988) (.814) (.890) (.868)

Concessions 4.86 4.72 5.33 4.82 4.43 5.14
 (1.33) (1.21) (1.20) (1.35) (1.48) (1.21)

Social Interaction 4.75 5.21 4.67 4.67 4.79 4.31 *
 (1.33) (1.33) (1.45) (1.45) (1.17) (1.24)

History 5.18 4.98 4.94 5.39 5.44 5.09
 (1.00) (1.09) (.861) (.862) (1.00) (1.11)

* p < .05 (significant when compared to control group)

EG1 = Experimental Group 1
EG2 = Experimental Group 2
EG3 = Experimental Group 3
EG4 = Experimental Group 4