ابعاد تقاضای بازار عمومی مرتبط با ورزش های تیمی حرفه ای : توسعه یک مقیاس

The purpose of this study was to develop the Scale of Market Demand to assess general market demand factors affecting the consumption of professional team sports, which was completed through the following five steps: (a) formulation of a theoretical framework, (b) development of a preliminary scale, (c) exploratory factor analysis (EFA), (d) confirmatory factor analysis (CFA), and (e) examination of predictive validity through conducting a structural equation modeling (SEM) analysis. Following a community intercept method, professional sport consumers (N = 453) in four southeastern metropolitan areas responded to the scale. Data were randomly split into two halves: one for EFA and the other for CFA. In the EFA with alpha extraction and promax rotation, six factors with 31 items emerged: opposing team, home team, game promotion, economic consideration, sport epitome, and schedule convenience. In the CFA with maximum likelihood estimation, five factors with 17 most pertinent items were retained, without the sport epitome factor. This five-factor model displayed good fit to the data, discriminant validity, and high reliability. The SEM revealed that home team, opposing team, and game promotion were predictive of game re-attendance behaviour.

According to Shank (2005), game attendance is the most traditional and important form of sport consumption behaviour. Chelladurai (1999) classified the sport industry into three segments: (a) sport economic activities, (b) spectator sports, and (c) participant sports. Of the three segments, spectator sports have been the fastest growing in terms of annual business transactions (Street & Smith's Sport Group, 2007). Other researchers have also recognised the continuous growth of spectator sports in North America and pointed out that spectator sports have become increasingly popular leisure behaviour for Americans (Ross and James, 2006 and Trail et al., 2005). The rapid and vast growth of professional sport teams in North America correlates highly with the increased interest in spectator sports among American people. Masteralexis, Barr, and Hums (2008) indicated that as of 2007, a total of 149 franchise teams belong to the five major professional sport leagues: Major League Baseball (MLB), National Football League (NFL), National Basketball Association (NBA), National Hockey League (NHL), and Major League Soccer (MLS). This figure does not include teams in less-prominent professional sport leagues such as the Women's National Basketball Association (WNBA), National Lacrosse League (NLL), and many other major and minor leagues. The augmentation of spectator sports has also been confirmed through attendance and media viewership rates. The same phenomenon is also true in NCAA Division I men's basketball and football, which are considered to be the two main revenue producers for collegiate athletic departments (Fulks, 2003). In 2005, nearly 1.3 million people watched March Madness men's college basketball games online (Rein, Kotler, & Shields, 2006). The increasing popularity of spectator sports has led to the establishments of new leagues (e.g., Women's Professional Soccer, National Lacrosse League), teams (e.g., Gold Coast Titans, San Jose Earthquakes, Oklahoma City Thunder), and multimedia outlets (e.g., NFL Sunday Ticket via Direct TV, ESPN 360), which have not only provided more spectating options for sport consumers but have also created greater competition among various leagues and teams for consumers. With such a crowded sport marketplace, sport consumers have many options in which to spend their leisure time and discretionary dollars. As a result, professional sport organizations face stiff competition in an effort to gain market share. Mullin, Hardy, and Sutton (2007) stated that “competition for sport dollar is growing at the pace of a full-court press” (p. 7) to describe the intensity of the competitive sport marketplace. According to Zhang, Smith, Pease, and Jambor (1997), market competition also comes from many other forms of leisure and entertainment options. As market competition becomes even more intense in professional sports, it is important for both researchers and practitioners to understand game consumption related variables in order to improve the quality of product offering and enhance competitiveness of the sport product (i.e., sport games). Also, marketers of professional sport teams need to develop strategic marketing plans that are based on an in-depth understanding of consumers. It is critical for them to identify those contributing variables that affect the consumption of spectators. Following an extensive literature review on factors influencing game attendance variables, Schofield (1983) proposed four market demand categories including demographic variables, economic variables, game attractiveness, and residual preference. Greenstein and Marcum (1981) and Jones (1984) focused their studies on game production functions and found that team performance variables, such as winning/losing record and presence of star player(s) were related to game attendance. Synthesizing key game demand variables and production functions, Zhang, Pease, Hui, and Thomas (1995) proposed the systematic concept of market demand, which was defined as the spectators’ expectations towards the main attributes of the core product (i.e., the game itself). Braunstein, Zhang, Trail, and Gibson (2005) further explained that market demand was a set of essential constructs associated with the game that a sport team could offer to its existing and prospective consumers. Essentially, it is a cluster of pull factors associated with the game that a professional sport team can offer to its prospective and returning spectators (Braunstein et al., 2005, Hansen and Gauthier, 1989 and Schofield, 1983). In a sense, market demand variables are comprised of the attributes of core game quality variables that are directly related to athlete/team performance, game schedule, and/or ticket affordability. Although market demand has consistently been found to be an important concept that influences sport consumption behaviour (Zhang et al., 2003b and Zhang et al., 1995), previous studies were usually conducted in the context of a specific sport (Braunstein et al., 2005, Zhang et al., 2003b and Zhang et al., 1995), lacking generalizability and application to a broader market environment of professional sports. To fill this void, the unique characteristics associated with general professional team sports were incorporated into the scale development in this study to inspire and enhance scientific inquiry into this topic area. It was anticipated that the developed Scale of Market Demand (SMD) would be frequently adopted by researchers and marketers to examine the target markets of professional team sports and their demands for the core attributes of game product. 1.1. Conceptual framework A theoretical justification for the market demand studies can be partially attributed to the cognitive-oriented attitude models, including the Adequacy Importance Model by Mazis, Ahtola, and Klippel (1975), the Theory of Reasoned Action proposed by Fishbein and Ajzen (1975), and the Theory of Planned Behaviour (Ajzen, 1991). The Adequacy Importance Model is a multi-attribute model that predicts consumer behaviour and was developed to understand consumers’ behaviour from a cognitive structure perspective. In this model, attitude was operationalized as an importance-weighted evaluation toward the attributes and dimensions of the products or services. Based on three studies, Mazis et al. (1975) found that this model explained consumer's behavioural intentions well. Mashiach (1980) supported the use of a multi-attribute model to predict sport consumers’ attendance behaviour by arguing that behavioural intentions formed by sport consumers were not a function of a single attribute, such as winning; instead, they were influenced by multiple attributes. In terms of predicting attitude in purchase behaviour, Oliver (1980) contended that attitude affects the purchase intentions of products and services. Consumers tend to form their attitude based on prior expectations regarding the performance of the product or service, which often results in ensuing behavioural intentions. The importance of attitude as a predictor of behavioural intentions was well explained by the Theory of Reasoned Action (Fishbein & Ajzen, 1975) and its updated version, termed the Theory of Planned Behaviour (Ajzen, 1991). These theories postulate that human behaviour is a direct consequence of behavioural intentions, which are functions of attitude, subjective norms, and perceived behavioural control. Several researchers have found that the attitude construct better explained behavioural intentions than subjective norms and perceived behavioural controls did (Stutzman and Green, 1982 and Warshaw et al., 1986). A strong attitude towards a certain object or phenomenon could act as a powerful heuristic that positively influences consumer behaviour (Fazio, Powell, & Williams, 1989). Likewise, when a sport consumer holds a positive attitude towards the attributes of game product such as home team/athlete performance, and/or game schedule, the positive attitude tends to be transformed into attendance and re-attendance behaviour (Zhang et al., 2003). Based on previous studies, it can be suggested that a multi-attribute model of market demand which reflects the unique nature of sport games can be a driving force of sport consumption behaviour, including game attendance or re-attendance. Following the various attitude models (i.e., Adequacy Importance Model, Theory of Reasoned Action, and Theory of Planned Behaviour) as primary theoretical frameworks and also based on the empirical findings of previous market demand studies (e.g., Braunstein et al., 2005, Zhang et al., 2004, Zhang et al., 2003b and Zhang et al., 1995), a six-factor model of general market demand was proposed in the current study. The six factors were: home team, opposing team, game promotion, economic consideration, sport epitome, and schedule convenience. Based on a review of related literature, the documentation and rationale for selecting each of the factors are presented below. 1.1.1. Home team The first proposed dimension, home team, is defined as the perceived quality of the home team that is represented by such attributes as home team performance, presence of superstar(s), athletic quality of home team players, win/loss record, home team reputation, and/or home team league standing. Previous studies found that this factor was most influential on the attendance of NBA (Zhang et al., 1995) and minor league hockey (Zhang et al., 1997) games. Zhang, Wall, and Smith (2000) found that win/loss record was positively related to NBA season ticket holder's game attendance. Bird (1982) in his football game attendance study found that league standing had a direct relationship with game attendance. 1.1.2. Opposing team Opposing team refers to the perceived quality of the opposing team and includes variables such as opposing team performance, athletic quality of opposing team, overall athletic quality of opposing team players, opposing team history and tradition, opposing team league standing, opposing team as a rivalry, and/or opposing team superstar(s). Madrigal (1995) found that competitive quality of opponent was related to affective reactions (Enjoyment and Basking-In-Reflected-Glory), both of which had a direct relationship with spectator satisfaction. In the context of the NHL, Jones (1984) found that the presence of star players on the home team, visiting team, or both was what motivated sport consumers to attend hockey games. Athletic quality of opposing team, opposing team history, league standing, and presence of superstar(s) on the opposing team were consistently found to be contributing variables to game attendance (Greenwell et al., 2002, Zhang et al., 1995 and Zhang et al., 1997). 1.1.3. Sport epitome Sport epitome is defined as the overall perception toward the general features of professional team sports. A wide range of sport epitome attributes have been identified as influencing variables, including but not limited to, closeness of competition, popularity of professional team sport, duration of game, high level of skills, best players in a sport, and speed of game. At times, this factor has been referred as ‘Love of a Sport’ (Braunstein et al., 2005, Ferreira and Armstrong, 2004 and Zhang et al., 2003b). Braunstein et al. found that this factor was an important factor representing Spectator Decision Making Inventory-Spring Training (SDMI-ST). Similar findings were discovered in Zhang et al.’s (2003) study on general market demand associated with professional sport consumption as this factor was found to be positively related to game attendance and media consumption. When studying game attributes that influenced college students’ game attendance, Ferreira and Armstrong (2004) found such variables as the duration of event, popularity of sport, high level of skill displayed, and speed of game were salient in influencing game attendance. Additionally, a number of previous studies on spectator motivation have recognised various sport epitome elements that were related to sport consumption. For example, closeness of competition was regularly identified as a variable for the drama dimension by Funk, Mahony, Nakazawa, and Hirakawa (2001), Trail and James (2001), and Wann (1995). Funk et al. and Trail and James also included high level of skills as a part of the physical skills of the athletes factor in their studies. 1.1.4. Economic consideration Economic consideration is defined as an individual's concern of economic issues, including such variables as ticket price, ticket affordability, good seats, and/or ticket discounts. Previous studies have shown statistical significance but conflicting results regarding the impact of economic consideration on game attendance. In their longitudinal study on MLB game attendance, Baade and Tiehen (1990) found that economic consideration was negatively related to game attendance. Similar findings were found in Bird's (1982) football game study and another study on general market demand of professional sports conducted by Hansen and Gauthier (1989). On the other hand, Zhang et al. (1995) found that ticket discounts, good seats, and group ticket cost were positively associated with the attendance of NBA games. 1.1.5. Game promotion Game promotion is defined as the specific mixture of marketing tools used by a sport organization for persuading consumers to attend the events or consume the product (Kotler & Armstrong, 1996). Fullerton and Merz (2008) argued that one of the frequently used marketing tools by professional sport marketers is sales promotion because oftentimes it helps to gain the attention of consumers, increase their interest, and eventually sell sport products. For instance, Pavelchak, Antil, and Munch (1988) found that people who recalled Super Bowl advertisements tended to possess a favorable attitude toward the companies and brands that were featured in commercials during the Super Bowl. Direct marketing has been found to be an effective promotional strategy that forms a positive relationship with the company (Fullerton and Merz, 2008 and Kotler and Armstrong, 1996). Recently, Seo and Green (2008) recognised that the Internet has become a crucial promotional tool used to influence sport consumption of professional sport consumers. The game promotion factor can be represented by such attributes as advertising, direct mail and notification, publicity, and web information. This factor should be distinguished from in-game entertainment amenities that can be manipulated by team marketers on a game-by-game basis (Zhang et al., 2005). Previous studies have indicated that game promotion activities (e.g., advertising, direct mail/notification) were positively related to various professional sports’ event attendance (Zhang et al., 2003b and Zhang et al., 1995). 1.1.6. Schedule convenience The last dimension in the proposed model, schedule convenience, is defined as the assigned time and day in which a sport game takes place. Zhang et al. (1995) found that schedule convenience was related to game attendance. Hill, Madura, and Zuber (1982) found that schedule convenience was positively related to game attendance at weekend and season ending games, but not afternoon games, in the MLB. Zhang (1998) examined minor league hockey spectators’ preferred time for game attendance and found that spectators preferred evening times (7:00 pm) for weekday and Saturday games, and late afternoon times (4:00 pm) for Sunday games. This factor appears to be less predictive of game attendance when compared to other market demand factors in the proposed model. However, in various studies on sport market demand, the schedule convenience factor usually emerged as an influencing dimension of consumer behaviour (Braunstein et al., 2005, Zhang et al., 2003b and Zhang et al., 1995). Numerous studies on sport market demand have been conducted to examine its ability to predict game attendance and/or re-attendance (e.g., Hansen and Gauthier, 1989, Zhang et al., 2004, Zhang et al., 2003b and Zhang et al., 1995). Market demand factors have been consistently identified as influencing factors for spectator attendance of professional team sport events (Braunstein et al., 2005, Greenstein and Marcum, 1981, Hansen and Gauthier, 1989, Jones, 1984, Zhang et al., 2003b and Zhang et al., 1995). Involving a sample of spectators of NBA regular season games, Zhang et al. (1995) found that four factors (home team, opposing team, game promotion, and schedule convenience) were related to NBA game attendance, with 6% of the variance explained. Similarly, Zhang et al. (2004) found that game attractiveness, economic consideration, and marketing promotion factors were positively related to past, current, and future consumption of NFL games, with 14% of the variance explained. In Zhang et al.’s (2003), study on general market demand of professional sports, Game Attractiveness and economic consideration factors were found to be predictive of game consumption level, with 22% of the variance explained. Although tremendous research efforts have been made, two major limitations have been identified in previous studies. First, previous studies mainly focused on specific professional sports such as the NBA (Zhang et al., 1995), MLB (Braunstein et al., 2005), and NFL (Zhang et al., 2004), suggesting a possible lack of a greater generalizability of the measurement instruments and the research findings to consumers beyond a specific sport. Second, Zhang, Lam et al.’s study was limited by including only three general factors in the study (i.e., game attractiveness, marketing promotion, and economic consideration) and ignored the presence of other possible factors. For instance, Braunstein et al. (2005) identified love of professional sport as an important dimension of market demand that was related to MLB Spring Training games. Schedule convenience is another factor that was consistently found to be an important factor of market demand. From an analytical perspective, Bagozzi (1980) argued that one reason for model misspecification in marketing research could be due to omitting important variables from the model. Further examination of the concept and factor structure of general market demand appears necessary in order to formulate general promotion guidelines for professional team sport researchers and practitioners. Therefore, the purpose of this study was to examine the dimensions of general market demand associated with professional team sports and develop a scale that reflected the identified dimensions. The objective of this study was accomplished through a comprehensive measurement process, including formulation of a preliminary scale, a test of content validity, a pilot study, test administration, and various statistical analyses of the scale's psychometric properties.

As market competition in professional team sports becomes more intense, it is imperative for both researchers and practitioners to identify those variables that directly and indirectly influence game consumption (Hansen and Gauthier, 1989 and Zhang et al., 1995). In-depth understanding of what factors influence spectators’ decision making to attend games, and how they refer the game products and services to others, is crucial for professional teams to better understand spectator consumption behaviour. Findings of previous studies revealed that market demand variables were important predictors of sport spectator consumption behaviour (Murray and Howat, 2002, Zhang, 1998, Zhang et al., 2004 and Zhang et al., 1995). Although previous researchers recognised the importance of market demand variables when marketing professional sport games, most of the studies were conducted in the context of a specific professional sport (Greenwell et al., 2002, Zhang et al., 2003a and Zhang et al., 1995). In these studies, scales developed under specific settings might not be directly applicable to other settings due to the lack of consideration of marketing characteristics that are generalizable to various professional team sport events. Thus, it was necessary to conduct this study that incorporated the attributes of core game products and market environment related to professional team sports (Mullin et al., 2007 and Zhang et al., 2003b). The current study was designed to develop the SMD measuring general market demand of professional team sports. In this study, psychometric testing procedures were meticulously conducted. For instance, systematic procedures were undertaken to formulate the preliminary scale, which included a review of literature, test of content validity by a panel of experts, and a pilot study by a group of consumers representing the targeted population. Previous research typically studied professional sport consumers at a limited number of sport events located in one geographic location. It was the intention of this study to include consumers of more diverse backgrounds in terms of geographic locations, sport types, consumption levels, and sociodemographics in an effort to improve the external validity of the developed scale. In this study, both EFA and CFA were conducted to ensure factor validity and generalizability of the resolved factor structure. For the SMD variables, six factors with 31 items were retained in the EFA: (a) opposing team, (b) home team, (c) game promotion, (d) economic consideration, (e) sport epitome, and (f) schedule convenience. The derived factors from the EFA were consistent with the theoretical dimensions suggested by previous researchers (Greenstein and Marcum, 1981, Hansen and Gauthier, 1989, Schofield, 1983 and Zhang et al., 1995). However, the six-factor model did not fit the data well in the initial CFA. After careful consideration of modification indices, the scale was revised to a five-factor model with a total of 17 items: opposing team (5 items), home team (3 items), game promotion (3 items), economic consideration (3 items), and schedule convenience (3 items). This respecified model exhibited much improved fit indexes. As a result of the respecification, the sport epitome factor was eliminated, mainly due to low indicator loadings. In previous studies, this factor was sometimes labeled as ‘love of the sport,’ and it was found to be a contributing variable to game attendance of intercollegiate competitions (Ferreira & Armstrong, 2004) and professional games (Braunstein et al., 2005, Zhang et al., 2003a and Zhang et al., 2003b). In Braunstein et al.’s (2005) study, the researchers found that Love of Baseball was an important factor related to MLB spring training game market demand; however, the factor in fact displayed poor psychometric properties. The factor was eventually retained by Braunstein et al. based on the consideration that Love of Baseball covered essential characteristics of MLB games, such as closeness of competition, duration of game, high level of skills, best players in a sport, and speed of game. Although the researchers of that study were reluctant to eliminate this factor, they did suggest the need for further studies examining this factor. The researchers of the current study conducted rigorous procedures in item purification, test of content validity, and a pilot study; yet, similar findings occurred with this factor. There was a wide array of aspects that were included in the sport epitome factor. Including a player perspective in this factor could have also led to correlations with the player aspects that were already in the previous dimensions. As Braunstein et al. suggested, additional examination of this factor is necessary in future studies. A key issue in future studies is how to keep the sport epitome factor theoretically separated from home team and opposing team factors because the factor analyses in the current study revealed that sport epitome items were double loaded on these two factors. Although the number of factors was reduced to five, the resolved constructs of the SMD were essentially consistent with previously suggested factors (Braunstein et al., 2005, Schofield, 1983, Zhang et al., 2004, Zhang et al., 2003a, Zhang et al., 2003b and Zhang et al., 1995). Nevertheless, findings of this study were likely more reliable and generalizable when considering the following aspects: (a) a more representative sample was involved and (b) more appropriate items were derived from advanced statistical analyses (i.e., CFA and SEM), In previous studies, data were usually collected on-site in arenas or stadiums (e.g., Zhang et al., 2003a and Zhang et al., 1995), where only spectators of one sporting event participated in the study and they might have been under temporal influence due to occurrence of winning or losing. The respondents of the current study were current sport consumers who indicated that they had attended one or more professional team sport events within the previous 12 months. Descriptive statistics indicated that a total of six premier professional team sport leagues were attended by the respondents, which may help improve the external validity of the findings and usefulness of the scale. Additionally, Zhang et al. (2003) suggested that besides EFA and CFA, other types of construct validity including convergent and discriminant validity tests be utilized to improve the factor structure. These suggestions were materialized in the current study, however, convergent validity of the current scale needs to be further validated in future studies. The current study retained at least three items per factor through the CFA. One of the limitations found in Zhang et al. (2003) study was that opposing team and schedule convenience factors were measured by only two items. When using CFA and SEM analyses, having the optimal number of items per factor is an important consideration for measurement precision (Bollen, 1989). The possible reason that two items of opposing team and schedule convenience factors were consistently used in previous studies may be due to the use of EFA as the primary item selection method, which is data-driven. Although our results were also data-driven due to reliance on modification indices when the scale was modified, the current study improved in that five items and three items related to opposing team and schedule convenience, respectively, were retained in the factor analyses (i.e., EFA and CFA). The items of the opposing team represented overall performance, athletic quality of the opposing team, athletic quality of players, opposing team's exciting play, and team reputation. Schedule convenience was represented by such attributes as game time of the day, convenient game schedule, and day of the week. In the current study, one of the items (i.e., home team reputation) in the home team factor did not meet the Anderson and Gerbing (1988) conservative criterion, but the item was retained due to its high level of theoretical relevance. Nonetheless, more research to validate the items related to the opposing team, schedule convenience, and home team factors appears necessary in future studies. Overall, the five-factor SMD measurement model displayed good psychometric properties, particularly in the areas of construct reliability and discriminant validity as verified by findings of EFA and CFA. Meanwhile, the research findings on content validity were still preliminary; yet, they showed promising results. The panel members for the test of content validity included academicians, practitioners, and sport consumers, who examined item relevance, representativeness, and clarity of the SMD items. Even so, a substantial number of initial items in the preliminary scale had to be excluded from the final version of the scale, suggesting that discrepancies existed between the expert panel and information obtained from the sport consumers. Future studies may consider applying a comparatively more rigorous approach of testing content validity in the social science domain as outlined by Aiken (1996) and Dunn, Bouffard, and Rogers (1999). In this contemporary approach, item examinations by the panel members are focused on the constructs. The three aspects of content validity (i.e., relevance, representativeness, and clarity) are simultaneously evaluated through various statistical procedures, including content validity coefficient, rater homogeneity coefficient, and rater reliability coefficient. Through this process, discrepant raters can be determined, more stringent items are retained, and weak or double loaded items are identified. Improved content validity would lead to improved convergent validity. Of the research findings, the high indicator loadings, high AVE values, and low interfactor correlations provided strong evidence for achieving a good factor structure of the SMD scale; nonetheless, no effort was made in this study to examine the convergent validity of the SMD factors. Convergent validity can usually be examined by relating to the scale and its factors of a previously validated measure that was developed in same or similar measurement environment and purpose. Although some similar factors were found in previous studies that measured market demand for professional sports (Zhang et al., 2003b and Zhang et al., 1995), they were developed in different marketing contexts. Thus, due to the absence of previously validated scales, no attempt was made in this study to adopt this approach. However, future studies should make an effort to accomplish this needed measurement aspect. Ideally, the multi-trait and multi-method approach should be adopted to substantiate and confirm the construct validity of the scale (Campell & Fiske, 1959). By implementing this advanced procedure, both convergent and discriminant validity can be examined simultaneously and a better understanding can be obtained not only on what is being measured, but also on how an attribute is being measured. A number of previous studies have provided evidence of criterion validity (e.g., sport consumption) for the market demand factors and indicated that market demand factors would explain 6–22% of game consumption variance (Zhang et al., 2003b and Zhang et al., 1995). In the current study, examination of criterion validity focused on one aspect of criterion validity, the predictability of the SMD factors to re-attendance intentions. Findings that the home team, opposing team, and game promotion factors were predictive of re-attendance intentions revealed that these factors were of basic capacity for analysing an individual's conative consumption behaviours, and these findings were consistent with those of previous studies (e.g., Braunstein et al., 2005, Hansen and Gauthier, 1989 and Zhang et al., 2003b). It is necessary to note that cumulatively, the three factors explained less than 17% of variance in re-attendance intentions, indicating that 83% was explained by concepts and factors not included in the SMD. In particular, none of the factors explained more than 8% of variance by itself. In addition, two other SMD factors (economic consideration and schedule convenience) explained no variance at all. Thus, the overall predictive validity was deemed partial. The negative relationship between game promotion and re-attendance intentions was interesting and somewhat different from the findings of previous studies. This finding may have in fact led to the following two speculations on the relevance level of game promotion to the market demand of professional team sports: (a) those with high demand for game promotion activities may have lower intention to attend sport events when compared to those with low demand for game promotion activities and (b) satisfying market demand for game promotion activities would likely be effective in making the game attractive to consumers with comparatively low consumption levels. Certainly, these speculations need further examination in future studies. The current study did not make an attempt to examine the predictability of the SMD factors to other past, present, and future consumption variables. This study also failed to examine the relationships between the SMD factors and those same or similar factors identified in previous studies (e.g., Braunstein et al., 2005, Zhang et al., 2003b and Zhang et al., 1995) by adopting the previously validated-test approach of criterion validity. Conducting all these tests in future studies would certainly be helpful in understanding all aspects of the criterion validity of the SMD scale. This study was also limited due to its sample size. Due to the necessity of conducting the EFA, only one half of the sample (n = 222) was used for the CFA. Although Wetson and Gore (2006) suggested that a minimum sample size of 200 was adequate for a CFA, the small sample size might have negatively influenced model fit for the market demand construct in the current study (Cheung & Rensvold, 2002). According to MacCallum, Roznowski, and Necowitz (1992), any model respecification should have an additional independent sample for cross-validation for the respecified model in order to avoid capitalization on chance. In the current study, the respecified model was not validated by an independent sample. Therefore, a careful interpretation of the results is required due to the following three reasons: (a) the preliminary model was revised to improve the overall model fit, (b) generalizability of the revised model to the population of interest for the current study remains unknown, and (c) the revised model was data-driven and considered as plausible, pending further validation using an independent sample in future studies. In the current study, no effort was made to examine if differences existed between die-hard and fair-weather fans in terms of the factor structures of the market demand construct. According to Wann and Branscombe (1990), spectators can at least be categorized as die-hard and fair-weather fans based on their consumption levels and socio-motivations. Die-hard fans generally are of higher team identification, involvement, and consumption levels than fair-weather fans; and are likely to support a team when the team does not perform well (Heere and James, 2007 and Trail et al., 2003). Perhaps, die-hard fans pay more attention to core product attributes (e.g., win/loss, level of performance, and/or the presence of star players); whereas fair-weather fans pay more attention to peripheral attributes (e.g., venue, promotion, and/or entertainment). These speculations need further examination by assessing invariance issues with respect to the factor structure of the SMD. When doing this, a number of sociodemographic and psychological variables, such as gender, team identification, and consumer involvement level, may be incorporated into the study. Additionally, data in this study were collected through a community intervention approach; thus, research participants were those who attended professional team sport events in the past. Due to the decay of memory, some of the respondents might not have been able to provide their responses with specificity. Hence, future studies should also examine the invariance issues between on-site and recall settings. With continued improvement, the SMD has great potential for adoption by researchers and practitioners. When considering its reasonable numbers of items (i.e., 17 items), the scale can be easily administered during on-site events or in various community locations. The scale can be used as a marketing tool for professional team sport organizations to better understand their spectator consumption behaviours. It can also be utilized in conjunction with other theoretically related variables, including but not limited to, game support programs, team identification, and word-of-mouth to examine how they function together to enhance sport consumption behaviours.