Brand personality in sport: dimension analysis and general scale development.

Braunstein, Jessica R. ; Ross, Stephen D.

Introduction

Long before Reis and Trout (1969) stressed the importance of positioning in order to solidify the branding process in the mind of one's consumer, marketers focused on specific "character" traits when promoting their products. Whether perceived or developed, brand personality (BP) has been studied regarding its use and effectiveness for decades (e.g., de Chernatony, 2001; Keller, 2003). With its academic exploration stemming from Aaker's (1997) original five dimensions of brand personality (i.e., Sincerity, Excitement, Competence, Sophistication, Ruggedness), this phenomenon has provided marketers with the ability to examine marketing practices, finding that matching the characteristics of a brand with those of its endorsers and consumers tends to be the most effective (e.g., Kamins, 1990, Lynch & Schuler, 1994). Embracing many general marketing practices, sport marketers have begun to question and, in turn, dissect this concept on their own turf (i.e., brand personality in sport--BPS) due to the unique characteristics often associated with sport and sport products. As such, this study proposes to assess BPS and the factors (i.e., characteristics) that may, or may not, define sport brands.

Brand Personality

As a result of advertising campaigns, brands are often portrayed as having human characteristics. Although this concept had been described, it had not been systematically and empirically studied until Aaker's (1997) creation of the Brand Personality Scale (BP). After examining 309 candidate traits based on previous literature (psychology, marketing, and original qualitative research), the author ultimately looked at 114 traits over 37 brands. The final BP scale included 42 items/traits under five factors: Sincerity (e.g., honest, genuine), Excitement (e.g., daring, spirited), Competence (e.g., reliable, responsible), Sophistication (e.g., glamorous, charming), and Ruggedness (e.g., tough, strong). Although Govers and Schoormans' (2005) longitudinal study confirmed Aaker's findings, examining the influence of product personality on a consumer's preference over time, Austin, Siguaw, and Mattila (2003) argued that the BP scale was not generalizable to individual brands, as a result of the method of the study. Agreeing that a measurement study was necessary to fully understand the concept of brand personality, Austin et al. (2003) reexamined Aaker's BP scale to determine the validity of those findings. The authors found that, while the constructs were internally reliable, they did not have construct validity.

Although the operationalization and validity of Aaker's original constructs continues to be questioned, the importance of the concept remains stable, as it continues to be used and examined in general business practices, particularly as the importance of branding has been stressed across a variety of disciplines. This focus on branding, concluding with the perception of a strong image in the mind of one's consumers (Reis & Trout, 1969), is ultimately a producer's nudge towards a potential (or current) customer's consumption. With a goal of developing roots for the eventual use of commercial imagery (i.e., print, television, and internet advertising), it is important for marketers to focus on the first step (i.e., internal branding) prior to their projection of their brand's image. Specifically, while the producer is, and should be, focused on the consumer's perception of the fit of message with that of the organization's characteristics (Fink, Cunningham, & Kensicki, 2004; Ohanian, 1991), they must take the time to first understand the perception of their brand before they can launch an appropriate marketing campaign. As numerous researchers have found that fit, when appropriate, does influence a consumer's purchase intentions (Tripp, Jensen, & Carlson, 1994; Wansink & Ray, 2000), it is important to develop a strong image as a sound foundation for a congruent message.

As previously mentioned, BP has received a lot of attention, and many studies have sought to conceptualize the construct and develop valid and reliable measurements in the general marketing literature (Aaker, 1997; de Chernatony, 2001; Keller, 2003). As sport marketing has evolved, and has adopted many of the strategic tactics of general business, brand personality in sport (BPS) has become a hot topic in the sport management literature (Gladden & Funk, 2002; Gladden & Milne, 1999; Parent & Seguin, 2008). While the primary focus has been on the characteristics of athlete endorsers (Boyd & Shank, 2004; Braunstein & Zhang, 2005, 2007) and the conditions that facilitate the transfer of BPS between a brand and a sports event/team (Gwinner & Eaton, 1999; Musante, Milne, & McDonald, 1999), the conceptual development of these characteristics (i.e., personality) for sport brands has recently come into question (Gladden & Funk, 2002; Gladden & Milne, 1999; Parent & Seguin, 2008). In sport, BPS may be influenced by a variety of factors, including: packaging, distribution, communication strategies (Gwin & Gwin, 2003), consumer interaction with the brand (Nandan, 2005), and the logo and the success of the team (Gladden & Milne, 1999). As such, these concepts must all be taken into account when determining the factors that make up BPS and, ultimately, an organization or entity's marketing mix.

Purpose

Although general business literature is pertinent in the assessment of the sports industry, the business of sport is unique, and should be assessed as such. As previously mentioned, the primary focus of the majority of the research in this area has focused on the characteristics of athlete endorsers (i.e., the Match-Up Hypothesis, or the "fit" between the athlete, the product, and the consumer; e.g., Kahle & Homer, 1985; Kamins, 1990). In order to properly address this concept, it is necessary to take a "step back" and assess the classification and use of these characteristics drawn from various disciplines. Questions still arise as to whether these current BP conceptualizations are appropriate for sport. Additionally, it is still unclear if current BP measures provide academicians and practitioners the appropriate tools to assess their product and market. As such, the purpose of this study was to continue this line of inquiry in the conceptualization and operationalization of BPS. Specifically, this study sought to examine and adapt preexisting BP measurement tools in order to formulate a scale that will evaluate the unique characteristics and/or personalities of professional sport brands.

Method

Measurements of BPS were collected from 449 students affiliated with two universities. Two different universities in varying geographic locations were selected in order to collect diverse information regarding professional teams and account for regional differences, and improve potential generalizability. Both universities were located in a major metropolitan area that was home to a professional team from each of the traditional big four sports. Students enrolled in a number of sport management, kinesiology, and general business courses were offered the opportunity to volunteer as study participants. Student samples in sport branding research are often utilized as they are significant consumers of sport, and the use of this population is common in brand choice research (Biswas & Sherrell, 1993). Additionally, the sampling method used in the current study mirrors other published research examining brand personality in sport (Ross, 2008).

Instrument

Respondents were asked to first list any professional sport team at the top of the survey, and then rate the degree to which they perceived BPS descriptor items as accurately describing the professional team listed. The BPS descriptor items utilized in the current research included 84 unique characteristic terms reflecting potential dimensions of BP in sport. These items were developed through an extensive review of BP literature, and included personality characteristic items from a number of previously developed scales purported to assess BP.

The BPS dimensions incorporated into the instrument were derived from Aaker's (1997) original Brand Personality Scale, as well as other previous adaptations of Aaker's instrument for use in sport (Braunstein & Zhang, 2005, 2007; Musante et al., 1999; Tenser, 2004). All potential items were screened for overlap, leading to several characteristic items being eliminated due to redundancy (i.e., exact duplication). The final survey instrument included a total of 84 descriptor items, all measured on a seven-point scale, where 1 = 'Totally Disagree' and 7 = 'Totally Agree'.

Analysis

A randomized assignment, split-sample technique was utilized in the current research for the analysis and development of the BPS scale. Using the random split function in Statistical Package for the Social Sciences (SPSS), each of the 449 respondents was placed into one of two data files. The first data file (N=225) was utilized for the exploratory factor analysis (EFA), while the second data file (N=224) was utilized for the confirmatory factor analysis (CFA) procedure.

The EFA was conducted on the first data set (N=225) using SPSS in order to examine the factor structure and determine the most appropriate number of BPS dimensions for further assessment. Maximum likelihood extraction was used as the EFA method, while varimax rotation was employed to examine a more interpretable solution. Given that there were no a priori hypotheses regarding the specific number of factors that should emerge, a variety of criteria were used to decide on an appropriate number of factors to retain: the Kaiser criterion (Kaiser, 1970), the scree test (Zwick & Velicer, 1982), parallel analysis (Zwick & Velicer, 1982), and extent of interpretability (Fabrigar, Wegener, MacCallum, & Strahan, 1999).

The validation of the scale was accomplished by submitting the second set of collected data (N=224) to a CFA using Linear Structural Relations (LISREL) 8.54. The purpose of the CFA was to estimate the proposed model for the scale items and constructs that were discovered in the initial EFA. Kline (1998) suggested using multiple fit indices in order to generate adequate information as to assess the overall fit data. As such, multiple goodness-of-fit measures were utilized to assess the results of the CFA model estimation. Specifically, the goodness-of-fit measures used in the current study were root mean square error of approximation (RMSEA), the Tucker-Lewis index (TLI), and comparative fit index (CFI).

In terms of reliability, the most important concern is the consistency of items within a measure. The reliability estimates were measured using the Cronbach's alpha correlation coefficient and the average variance extracted (AVE). Cronbach's alpha coefficient is a test of internal consistency that tests the homogeneity of all the items in the instrument subscale. The variance explained by each of the identified constructs relative to the amount of variance attributed to measurement error (AVE) was also examined as a measure of reliability (Fornell & Larker, 1981).

In addition to evaluating the fit of the proposed model through the CFA and examining the reliability estimates, attempts were made to further establish construct validity through tests of discriminant validity and convergent validity. Convergent validity was assessed by examining each item's loading on the construct on which it loaded and the standard error for which it was associated (Anderson & Gerbing, 1988). Discriminant validity was assessed through two methods: examination of the correlations between constructs (Anderson & Gerbing, 1988) and evaluation of the AVE values (Fornell & Larker, 1981).

Results

Exploratory Factor Analysys (EFA)

The Kaiser criterion, assessing the number of factors with eigenvalues greater than 1.0, suggested retaining 18 factors. The scree test revealed a substantial drop in eigenvalues after six factors, while the parallel analysis suggested retaining six factors. In contexts in which procedures produce different numbers of factors, the researchers should examine the models to determine which is most plausible (Ford, McCallum, & Tait, 1986). The rotated solutions for these models can be examined to see which model produces the most readily interpretable and theoretically sensible pattern of results (Comrey, 1978). As such, the interpretability (Fabrigar et al., 1999) of factor loadings suggested retention of six factors.

The item loadings were assessed to ensure that the items loaded significantly on the respective factors. Particular attention was paid to the number of significant items loading on the respective factors to avoid retaining factors with only one item or too many items that could be reasonably interpreted (Fabrigar et al., 1999; Fava & Velicer, 1992). Interpretability of respective factors was also an important criterion in selecting the number of factors to retain. Given the potential hazards of overfactoring and underfactoring, in addition to the extent to which the solutions were interpretable, the six-factor solution was accepted. The resulting structure of the instrument included 40 items assessing the following six identified factors: Competence (14 items), Sophistication (10 items), Sincerity (7 items), Ruggedness (3 items), Community-driven (3 items), and Classic (3 items).

Confirmatory Factor Analysis (CFA)

The fit indices produced through the CFA indicated that the model provided adequate fit for the proposed factor dimensions. In the model examined here, the RMSEA, TLI, and CFI values are all above the acceptable levels. The RMSEA (0.071) was well below the maximum cut-off of 0.10 (Steiger, 1998), while both the TLI (0.95) and the CFI (0.96) indicated an adequate fit of the data given that each met the recommended criteria of 0.90 (Kline, 1998). Taken as a whole, the goodness of fit statistics produced by the CFA indicated that the overall data fit the model satisfactorily.

Reliability Estimates

Cronbach's alphas were calculated for each of the six dimensions utilizing their corresponding items as defined in the proposed model. The reliability of four of the six dimensions met the recommended criteria set by Nunnally and Bernstein (1994) (see Table 1), and ranged from 0.65 to 0.93. The Ruggedness and Community-driven dimensions failed to meet the .70 minimum alpha criteria. Fornell and Larcker (1981) suggest that the AVE of any construct should be greater than the unique variance. More specifically, the AVE values of each identified factor in a model should exceed a value of 0.50. The AVE values in the current study were calculated by averaging the squared multiple correlations derived from the reported standardized loadings. These values ranged from 0.33 to 0.57 (see Table 1), and revealed that four of the six dimensions failed to meet the recommended criteria. The Success and Classic factors were the only two BPS dimensions that satisfied the 0.50 minimum AVE criterion level.

Convergent and Discriminant Validity

A fundamental aspect of establishing construct validity is establishing convergent validity, and whether each of a proposed scale items contributes to its underlying theoretical construct. Specifically, if each indicator's loading on its posited underlying construct is greater than twice its standard error then convergent validity can be verified (Anderson & Gerbing, 1988). The results of the data analysis for the current study indicated that each of the items met this criterion (see Table 1). Often, the residual matrix produced by the CFA is examined as another measure of the internal quality of a construct's items. Hair, Black, Babin, Anderson, and Tatham (2005) suggest that when inspecting the residual matrix, the standardized values should not exceed a 2.58 absolute value. After reviewing the standardized residuals produced by the analysis in the current study, it was found that only a relatively small percentage of the standardized residuals (11.9%) surpass the recommended criterion put forth by Hair et al. (2005).

When assessing discriminant validity, according to Anderson and Gerbing (1988), the correlation between any two constructs should not be within two standard errors of unity. The results indicate that although some correlations between the constructs were quite high, none failed this initial test of discriminant validity (see Table 2). However, evaluation of each construct's AVE provides a more rigorous test of discriminant validity, and as Fornell and Larker (1981) suggest, the AVE for each construct should be greater than the squared correlation between that construct and any other. Failing to meet this criterion denotes a lack of discrimination. As shown in Table 2, some of the proposed factors correlate with factors from which they should differ, and thus indicates that the variance accounted for by these factors explain less variance than the factors with which it correlates.

Conclusions and Discussion

The findings suggest mixed results in terms of the fit of the newly adapted scale. While the CFA suggested the data adequately fit the model, the reliability and validity of the BPS scale, in its current state, do not provide a sufficiently sound instrument. However, we do believe that these findings (i.e., six factors with 41 items) provide a sound foundation for the further development of the theoretical framework of BPS.

Generally speaking, the estimates of reliability provided a mixed bag of results. The Success and Classic factors showed good reliability estimates in both the Cronbach's alpha and AVE. Conversely, the Rugged and Community-driven factors showed unreliability in both criteria. The Sophistication and Sincerity dimensions showed good alpha levels, but poor AVE estimates. These results are hypothesized to have occurred for a number of reasons. The primary reason for the incongruent findings is the number of specific items used to measure each of the factors. Specifically, the Success dimension was measured using 14 items, and found to be reliable on all criteria. However, the Community-driven dimension was only measured using three items and was found to be unreliable on both criteria. It is possible that the number of items to measure Community-driven was not enough and wide-ranging to fully measure the concept of involvement in the community. Additionally, when examining the specific items used to assess Community-driven, questions arise on whether the items truly reflect the concept. Specifically, the notions of authenticity (Alexander, 2008), inspiration (Motion, Leitch, & Brodie, 2003), and service-orientation (Gronroos, 1989) have been discussed when defining other constructs. One suggestion for future research would be to investigate if these items are truly linked conceptually to community involvement, and if so, then how to better reflect the idea when measuring the dimension.

In terms of validity assessment, the results prove to be more encouraging. While the examination of convergent validity results showed evidence of good convergent validity, the discriminant validity tests were not as conclusive. One criterion showed good results, while the other tests of discriminant validity showed problems with the measure. The majority of the issues of discriminant validity were based on the Community-driven factor. Specifically, there were fairly high correlations between Community-driven, and Success (.787), Sincerity (.723), and Classic (.729). While the conceptual linkage between some of these factors is logical, discriminant validity still needs to be addressed in future research. The idea of being involved and committed to the community can be very clearly linked to the idea of an organization being sincere. Therefore, the correlation between these two factors does make conceptual sense.

The results of this study show promising structures for the BPS model. However, it is clear that additional work must be competed to ensure a sound measurement tool for the future. It is suggested here that future research be conducted in order to heighten the discriminant validity of the measure. The conceptual linkage between dimensions needs to be investigated further in order to truly partial out what items comprise each factor. For example, future research could better explain and delineate between Community-driven and the Sincerity in order to, perhaps, find a common theme. Interestingly, the Excitement factor proposed by Aaker (1997) was not found in the current study. This is perhaps due to the specific items comprising the original Excitement dimension. Specifically, the items were related more to novelty issues rather than exhilaration or the thrill of sport. For example, items such as imaginative or up-to-date are not conceptually relevant to the excitement or drama associated with sport.

Additionally, it is suggested here that modifications be made to the respondent instructions in order to eliminate bias responses. Specifically, the respondents were instructed that the items were examples of personality traits. Utilizing this wording might inherently project a humanizing context onto the brand object, and thus have influenced the way in which participants in this study, and other BP studies, responded to the items. Perhaps making use of the term "characteristic" might remove some of this bias. However, caution must be used when establishing these participant instructions, as any radical changes might change the context in which the construct is conceptualized.

Managerial Implications

Brand personality is an important construct for managers to understand in order to effectively market and position a brand. That is, sport marketers could utilize collected brand personality information in the development of marketing and positioning strategies with their respective organizations. Specifically, sport organizations can utilize the BPS information as a point of differentiation for the brand, and perhaps to position the brand against competitors. It is vitally important for organizations to develop an understanding of competitors BPS in order to fully maximize marketing strategies. Too often organizations are narrowly focused on identifying images associated with their own organizations while ignoring the associations linked to the competition. Through a more complete understanding of both the home organization and that of competitors, sport teams could direct resources into promoting the BPS dimensions that are in most need of change or strengthening.

More fundamentally, sport teams can use the BPS tool developed in the current study to assess the current BPS of the organization, and then determine if specific characteristics need to be strengthened, augmented, or even deemphasized. In some instances there may be one specific BPS dimension that reflects well on the organization, and thus the organization would want to leave it untouched. Often, these characteristics are in line with the vision, mission, and values of the organization. In other instances, there may be a desired BPS dimension that is not evoked, and thus needs to be strengthened to maximize the full potential. Finally, it is possible that there may be situations in which an organization has a strong negative BPS dimension reflecting poorly in the brand. In such cases, the BPS dimensions needs to be deemphasized and augmented in order to relieve the negative connotation. In all of these situations, effective measurement of the BPS is crucial to properly manage and market the brand for success.

It is also suggested that this, and other BPS measurement tools, be utilized for sponsorship management purposes. A substantial amount of research has suggested the importance of image matching in sponsorship management (Gwinner, 1997; Musante et al., 1999), and such measures of BPS provide a way to assess a brand's personality prior to sponsorship contract agreements. That is, this tool could be used to assess the BPS of a brand (or potential sponsor) to determine if the sponsorship partnership is a good fit for the image of each organization. Furthermore, in terms of sport celebrity endorsements, BPS tools can provide academicians and practitioners with a scale through which the match-up hypothesis (Kahle & Homer, 1985) can be applied.

Future Research

As with any study, the findings of the current research point towards many topics for future examination. Perhaps the most interesting of the findings stems from the specific dimensions identified in the exploratory factor analysis, and the accompanying items that comprise each of those factors. Specifically, the results suggest that perhaps sport brands have a "layered" personality that could be managed and marketed on multiple levels. For example, the dimension of Success, and the individual items that make up the factor, seem to emit some of the images for the core product for sport organizations. For example, the traits of successful, superior, and quality are indicative of competitive sport in general, and critical elements of the sport product (Mullin, Hardy, & Sutton, 2007).

Additional "layers" of the brand personality might include the "management of the brand" where the factors of Sophistication and Community-driven might be located, and thus represent extended elements of the core brand (Mullin et al., 2007). Aaker (1996) suggests that a specific brand identity perspective is viewing the brand as an organization, of which organizational attributes such as sophistication and consumer concern are emphasized. Furthermore, there are intangible components of the extended brand that may make up additional layers of the sport team's BPS. Specifically, this research identified the factors of Classic and Sincerity as being dominant dimensions of the BPS construct. Interestingly, these dimensions are based on intangible consumer perceptions, and cannot be measured based upon preset criteria. That is, the Classic dimension represents the perception of a team as being traditional or old-school (Aiken & Sukhdial, 2004), while Sincerity can be described as fan-oriented or authentic. These dimensions are highly subjective and can often differ between teams and individuals.

The proposition of a layered BPS also suggests reconceptualization for the sport brand personality construct around the complete "brand" rather than just the team. That is, the "brand" for the sport organization includes much more than just the team that plays during competitions. There are multiple components for the sport brand and include, but are not limited to, the front-office personnel, brand extensions, and the facilities associated with the team. Perhaps it is necessary to take a few steps back, and begin the development of a sport brand personality scale by examining the construct in this manner. As Churchill (1979) proposes, the first step in scale development is the specification of the construct's domain. Through this process, a more complete understanding of all the elements that encompass a sport brand might be possible.

Final Thoughts

While branding has become an integral part of the marketing exchange process (i.e., producer to consumer and back again), it is only productive when the consumer buys into the entity's developed brand. If a consumer has a clear image of a brand, consistent with that of the producer, this could lead to a very clear marketing message. However, damage can be done when the consumer does not associate the brand with the characteristics that the producer is attempting to develop and, in turn, focusing marketing efforts on. As such, the BPS scale was developed in order to begin to analyze the effectiveness of the projected sport brand on the consumer.

We believe that the findings presented here are vital to the further development of both the theoretical framework and application of the BPS scale, providing a sound preliminary exploration of the operationalization of brand personality in sport. In addition to expanding upon current theories and applications in this area, this work subscribes to the belief that a scale for both practitioners and academicians can be used to develop a more effective marketing mix for individual sport entities. As the current economic climate continues to impact sport, and marketing budgets are slashed, the further development of tools that provide the opportunity for marketers to take a strategic approach will prove beneficial to academicians as well as practitioners.

References

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

Aaker, J. L. (1997). Dimensions of brand personality. Journal of Marketing Research, 34, 347-357.

Aiken, D., & Sukhdial, A. (2004). Exploring the old school concept: Adding definition to a "new" market segmentation dimension. Sport Marketing Quarterly, 13, 73-81.

Alexander, N. (2008). Brand authentication: Creating and maintaining brand auras. European Journal of Marketing, 43(3/4), 551-562.

Anderson, J., & Gerbing, D. (1988). Structural equation modelling in practice: A review and recommended two-step approach. Psychological Bulletin, 10, 411-423.

Austin, J., Siguaw, J., & Mattila (2003). An assessment of the Aaker brand personality framework: Method, measurement and conceptual issues. Journal of Strategic Marketing, 11(2), 77-92.

Azoulay, A., & Kapferer, J. (2003). Do brand personality scales really measure brand personality? Brand Management, 11(2), 143-155.

Biswas, A., & Sherrell, D. (1993). The influence of product knowledge and brand name on internal price standards and confidence. Psychology and Marketing, 46(1), 31-46.

Boyd, T. C., & Shank, M. D. (2004). Athletes as product endorsers: The effect of gender and product relatedness, Sport Marketing Quarterly, 13, 82-93.

Braunstein, J. R., & Zhang, J. J. (2005). Dimensions of athletic star power associated with generation y sport consumption. International Journal of Sports Marketing and Sponsorship, 6, 242-67.

Braunstein, J. R., & Zhang, J. J. (2007, May). Influence of product-endorser match-up on consumer's purchase intentions of (non-sport) endorsed products: Development of a model. Presented at the 2007 North American Society for Sport Management Conference. Ft. Lauderdale, FL: University of Miami.

Churchill, G. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(1), 64-73.

Comrey, A. (1978). Common methodological problems in factor analytic studies. Journal of Consulting and Clinical Psychology, 46, 648-659.

de Chernatony, L. (2001) A model for strategically building brands. Journal of Brand Management, 9(1), 32-44.

Fabrigar, L., Wegener, D., MacCallum, R., & Strahan, E. (1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4, 272-299.

Fava, J., & Velicer, W. (1992). The effects of overextraction on factor and component analysis. Multivariate Behavioral Research, 27, 387-415.

Fink, J. S., Cunningham, G. B., & Kensicki, L. J. (2004). Utilizing athletes as endorsers to sell women's sport: Attractiveness versus expertise. Journal of Sport Management, 18, 350-367.

Ford, J., McCallum, R., & Tait, M. (1986). The application of exploratory factor analysis in applied psychology: A critical review and analysis. Personnel Psychology, 39, 291-314.

Fornell, C., & Larcker, D. (1981) Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39-50.

Gladden, J. M., & Funk, D. C. (2002). Developing an understanding of brand associations in team sport: Empirical evidence from consumers of professional sport. Journal of Sport Management, 16(1), 54-81.

Gladden, J. M., & Milne, G. R. (1999). Examining the importance of brand equity in professional sport. Sport Marketing Quarterly, 8(1), 21-29.

Govers, P. C. M., & Schoormans, J. P. L. (2005). Product personality and its influence on consumer preference. Journal of Consumer Marketing, 22(4), 189-197.

Gronroos, C. (1989). Defining marketing: A market-oriented approach. European Journal of Marketing, 23(1), 52-60.

Gwin, C., & Gwin C. (2003). Product attributes model: A tool for evaluating brand positioning. Journal of Marketing Theory and Practice, 11(2), 30-42.

Gwinner, K. (1997). A model of image creation and image transfer in event sponsorship. International Marketing Review, 14(3), 145-158.

Gwinner, K. P., & Eaton, J. (1999). Building brand image through event sponsorship: The role of image transfer. Journal of Advertising, 28(4), 47-57.

Hair, J., Black, B., Babin, B., Anderson, R., & Tatham, R. (2005). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Prentice Hall.

Kahle, L. R., & Homer, P. M. (1985). Physical attractiveness of celebrity endorser: A social adaptation perspective. Journal of Consumer Research, 11, 954-961.

Kaiser, H. (1970). A second generation little jiffy. Psychometrika, 35(4), 401-415.

Kamins, M. A. (1990). An investigation into the match-up-hypothesis in celebrity advertising: When beauty be only skin deep. Journal of Advertising, 19(1), 4-13.

Keller, K. L. (2003). Strategic brand management: Building, measuring and managing brand equity (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall.

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

Lynch, J., & Schuler, D. (1994). The matchup effect of spokesperson and product congruency: A schema theory interpretation. Psychology and Marketing, 11, 417-445.

Motion, J., Leitch, S., & Brodie, R. (2003). Equity in corporate co-branding: The case of adidas and the All Blacks. European Journal of Marketing, 37(7/8), 1080-1094.

Mullin, B., Hardy, S., & Sutton, W. (2007). Sport Marketing (3rd ed.). Champaign, IL: Human Kinetics.

Musante, M., Milne, G. R., & McDonald, M. A. (1999). Sport sponsorship: Evaluating the sport and brand image match. International Journal of Sports Marketing and Sponsorship, 1(1), 32-37.

Nandan, S. (2005). An exploration of the brand identity-brand image linkage: A communications perspective. Brand Management, 12(4), 264278.

Nunnally, J., & Bernstein, I. (1994) Psychometric theory (3rd ed.) New York: McGraw Hill.

Ohanian, R. (1991). The impact of celebrity spokesperson's perceived image on consumer's intention to purchase. Journal of Advertising Research, 31(1), 46-52.

Parent, M., & Seguin, B. (2008). Toward a model of brand creation for international large-scale sporting events. Journal of Sport Management, 22, 526-549.

Ries, A., & Trout, J. (1969). Positioning: The battle for your mind. New York: McGraw-Hill.

Ross, S. (2008). Assessing the use of the brand personality scale in team sport. International Journal of Sport Management and Marketing, 3(1/2), 23-38.

Steiger, J. (1998). A note on multiple sample extensions of the RMSEA fit index. Structural Equation Modeling, 5, 411-419.

Tenser, J. (2004). Endorser qualities count more than ever. Advertising Age, 75(45), S2-S4.

Tripp, C., Jensen, T. D., & Carlson, L. (1994). The effects of multiple product endorsements by celebrities on consumers' attitudes and intentions. Journal of Consumer Research, 20, 535-547.

Wansink, B., & Ray, M. L. (2000). Estimating an advertisement's impact on one's consumption of a brand. Journal of Advertising Research, 40(6), 106-113.

Zwick, W., & Velicer, W. (1982). Factors influencing four rules for determining the number of components to retain. Multivariate Behavioral Research, 17, 253-269.

Jessica R. Braunstein, PhD, is an assistant professor and the sport management internship coordinator at Towson University. Her research interests focus on consumer behavior in sport, specifically examining the use and effectiveness of athlete endorsers, brand personality, and the internal and external motivating factors regarding spectator consumption. 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.

Factor Reliabilities, AVE, Item Loadings, Standard Errors, and
T-values for the BPS Dimensions

Item [alpha] AVE Factor-Loading

Success .93 .50
 1. successful .754
 2. Efficient .765
 3. high-performance .711
 4. dependable .719
 5. superior .730
 6. accomplished .711
 7. respected .765
 8. reliable .735
 9. confident .676
 10. quality .692
 11. consistent .612
 12. capable .622
 13. mature .732
 14. hard-working .621

Sophistication .83 .33
 1. stylish .527
 2. up-to-date .554
 3. appearance .592
 4. glamorous .519
 5. flashy .595
 6. trendy .670
 7. upper class .668
 8. sophisticated .533
 9. attractive .609
 10. corporate .395

Sincerity .86 .47
 1. honest .734
 2. genuine .672
 3. sincere .659
 4. down-to-earth .700
 5. charming .669
 6. friendly .619
 7. family-oriented .724

Rugged .65 .41
 1. bold .789
 2. daring .678
 3. rugged .369

Community-driven .68 .42
 1. authentic .732
 2. inspirational .669
 3. service-oriented .520

Classic .80 .57
 1. traditional .648
 2. classic .803
 3. old fashioned .805

Item Standard Error t

Success
 1. successful .057 13.06
 2. Efficient .057 13.33
 3. high-performance .059 12.05
 4. dependable .058 12.23
 5. superior .058 12.48
 6. accomplished .058 12.03
 7. respected .059 13.34
 8. reliable .058 12.61
 9. confident .060 11.26
 10. quality .059 11.61
 11. consistent .061 9.90
 12. capable .061 10.12
 13. mature .058 12.53
 14. hard-working .061 10.10

Sophistication
 1. stylish .066 7.94
 2. up-to-date .065 8.42
 3. appearance .064 9.12
 4. glamorous .066 7.81
 5. flashy .065 9.18
 6. trendy .062 10.66
 7. upper class .062 10.63
 8. sophisticated .066 8.05
 9. attractive .064 9.46
 10. corporate .068 12.21

Sincerity
 1. honest .060 12.21
 2. genuine .061 10.86
 3. sincere .062 10.8
 4. down-to-earth .061 11.46
 5. charming .062 10.78
 6. friendly .063 9.76
 7. family-oriented .060 12.00

Rugged
 1. bold .071 11.08
 2. daring .070 9.60
 3. rugged .073 4.99

Community-driven
 1. authentic .061 11.98
 2. inspirational .062 10.75
 3. service-oriented .064 8.01

Classic
 1. traditional .064 10.02
 2. classic .061 13.16
 3. old fashioned .061 13.20

Table 2.

Correlations Between Factors and Standard Errors (in parentheses) for
the BPS Confirmation Model

Factors Success Sophistication Sincerity Rugged

Success 1.000
Sophistication .716* 1.000
 (.041)
Sincerity .677 .539 1.000
 (.044) (.059)
Rugged .545 .607* .348 1.000
 (.063) (.063) (.077)
Community-driven .863* .694* .880* .454
 (.038) (.058) (.041) (.081)
Classic .528 .397 .542 .147
 (.058) (.070) (.060) (.085)

Factors Community-driven Classic

Success
Sophistication

Sincerity

Rugged

Community-driven 1.000

Classic .779* 1.000
 (.052)

Note: * denotes failing the second discriminant validity test