Ticket Price Comparison of Double-A and Triple-A Affiliate Baseball Leagues

### Abstract

As the economy continues to decline, sport managers realize that discretionary spending is limited. As such, sport managers are giving more consideration to price strategies within their own marketing mix as well as their comparison to other sport teams. The purpose of this study was to conduct a cross-sectional pricing investigation of individual teams by region within a Class-AAA and Class-AA league from the minor league baseball system. Data were obtained for ticket prices and fees from baseball team websites and phone interviews. Multivariate analysis of variance was examined for both Double-A and Triple-A leagues divided into regions. This study found no significant F (1,6) = .09, p = .77 differences for the highest ticket prices, F (1,6) = .09, p = .78, or the lowest ticket prices, and F (1,6) = .07, p = .80 for the groups within the Double-A Affiliate Texas League. However, a significance F (2,13) = 8.08, p = .00 was found in lowest ticket price within the Triple-A Affiliate Pacific Coast League, unlike highest ticket prices and fees which were not measurably different. Most minor league sport managers could consider this advantageous for promoting their entertainment as a good economic value.

**Key Words:** Baseball, Ticket Prices, Minor League

### Introduction

In light of recent economic times, sport organizations are faced with the challenge of maintaining a competitive marketplace while keeping a close eye on the bottom line. At the same time, the economic market watch (9) indicates consumers are becoming more selective with discretionary spending. Since sport consumption is not a fundamental cost of living, sport organizations have had to take a hard look at their strategic placement in the market. Pricing is a fundamental component of the marketing mix (6). Economic strategists have recommended complex formulas to establish pricing structures (2), while many sport organizations are opting for simplicity (6). In fact, Mullin, Hardy, and Sutton (6) said “the core issues in any pricing situation are cost, value, and objectives” (p. 215). In keeping with the simple pricing strategies that are the focus today, many sport franchises have utilized price comparisons as a simple and effective method of determining where a sport organization “fits” in the regional and league markets.

#### Price Comparisons

Determining the best fit in the sport consumer marketplace and how pricing strategies align with peer teams has become an emphasis for sport managers within the minor league baseball industry. As baseball ticket prices increase (1) and pricing strategies become complex (7), baseball consumers may look to alternative discretionary spending investments. Price comparison consumption behaviors have increased exponentially with the convenience of the Internet (5). Websites like Pricerunner, Amazon, and Shop.com have allowed potential consumers to price shop merchandise with several companies at the same time. Sport organizations have not considered price comparisons as a major influence on strategic pricing, due to the uniqueness offered in sport consumption. For example, sporting events occur sometimes great distances apart whereby a potential consumer may traditionally only be willing to travel 30 miles (4) and therefore do not offer a competitive risk to the local sport organization. However, as seen in recent articles (e.g., 3, 8) with a click of a button, price comparisons are made. In today’s tumultuous economy, many sport organizations have elected to market their event as a “value” within the discretionary spending category. This marketing technique is not only being utilized in relation to their direct sport competition, but also with discretionary spending businesses in general (e.g. cinema, concerts, other types of sporting events).

It was hypothesized that there was a significant (p < .05) difference between Texas and Non-Texas regions when comparing ticket pricing (highest price, lowest price, and ticketing fee) for minor league baseball Class-AA Texas League, as well as a significant (p < .05) difference among West, South, and Central regions when comparing ticket pricing (highest price, lowest price, and ticketing fee) for minor league baseball Class-AAA Pacific Coast League. Additionally, it was hypothesized that there was a significant (p < .05) difference between Double-A and Triple-A affiliate leagues when comparing ticket pricing (highest price, lowest price, and ticketing fee). This study examined price comparisons of minor league professional baseball teams segmented by league and region. A selection criterion was based upon geographic region of the minor league baseball teams as well as a comparison between Class-AA and Class-AAA organizations. Tables 1 and 2 represent the teams included in the study organized by league and region.

### Methods

#### Procedures

The data were collected through a variety of methods. Most of the information was collected through individual team websites. Some information was obtained though cold-calling via landline phones, and remaining data were provided through personal interviews. Once the data were collected, they were entered into a Microsoft Excel spreadsheet and reserved for future reference. Collection of data occurred over several months during the 2009 baseball season.

#### Data Analyses

Descriptive statistics, specifically means and standard deviations, were initially reviewed and reported for the Leagues, regions, and individual teams. The data obtained for the purpose of determining the research hypotheses were analyzed using MANOVA statistical methods. The independent variables were regions within the Texas League (Texas and Non-Texas regions) and Pacific Coast League (West, South, and Central regions). The dependent variables were the ticket pricing (highest ticket price, lowest ticket price, and ticketing fee), concession pricing (draft beer and hot dogs), and price for a family of four. Data were analyzed using Statistical Package for the Social Sciences (SPSS) version 17.0.

### Results And Discussion

#### Descriptive Statistics

Descriptive statistics were reported on all the dependent variables for the team, region and league. Table 3 provides the means and standard deviations found for highest ticket price, lowest ticket price, and ticketing fees.

Further examination of the ticket pricing established by the respective franchises indicates that the Triple-A teams (i.e. Pacific Coast League) have the greatest price point means. Specifically, the West region is higher than all other regions examined across all three price point values $22.63, $7.56, and $3.66 respectively. Inversely, the Southern region within the same league offers considerably lower price points for the highest ticket ($12.00) and lowest ticket ($5.50). All regions studied included fees into their ticket prices, particularly when utilizing online purchasing websites. Most regions were consistently adding approximately $2.00 to the overall price of the ticket.

Figures 1 and 2 show graphical comparisons between the highest and lowest ticket prices for the Texas League teams and Pacific Coast League teams respectively. As shown in both the Texas League and the Pacific Coast League price comparisons, there is great variability among teams when comparing the highest ticket prices; however within both leagues all the franchises have a relatively similar low cost for tickets.

The aforementioned price points did not include additional fees traditionally included in ticket prices for sporting events. As an example of how prices fluctuate with fees included in the price, Figures 3 and 4 show the price increase for the highest ticket price per franchise within both the Texas League and Pacific Coast League. Previously noted within Table 3, the West region of the Pacific Coast League had the highest ticket prices and once again that is reflected in Figure 4 as the fees are also the greatest among several West coast baseball franchises.

#### MANOVA Hypotheses Testing

The three hypotheses were tested by applying MANOVA to the data with SPSS software. The first group analyzed was the Texas League regions (Texas and Non-Texas) as the independent variable and the ticket prices (highest, lowest, and fees) as the dependent variables. As indicated in Table 4, there were no significant F (1,6) = .09, p = .77 differences for the highest ticket prices, F (1,6) = .09, p = .78, and the lowest ticket prices, F (1,6) = .07, p = .80, or the ticketing fees between Texas teams and Non-Texas teams within the Double-A Affiliate Texas League baseball.

As shown in Table 5, the same statistical principles were applied to the Pacific Coast League. The three regions, West, South, and Central, were the independent variables and the ticket prices (highest, lowest, and fees) were the dependent variables. There was a significant F (2,13) = 8.08, p = .00 difference in lowest ticket price between Pacific Coast League when divided by region. A Scheffe post hoc analysis revealed that the South region was significantly different from both the West (p = .00) and the Central (p = .04) regions. The South region had the lowest of the low ticket prices with an average of $5.50 as compared to the West which was $7.56 and the Central at $7.25.

Table 6 shows the difference between the Double-A Texas League and the Triple-A Pacific Coast League ANOVA source table. As noted in the Table, there were no significant F (1,22) = 1.91, p = .18 differences for the highest ticket prices, F (1,22) = 4.11, p = .06, the lowest ticket prices, and F (1,22) = .66, p = .42, or the ticketing fees.

### Conclusions

As more sport franchises compete in this challenging economic market, the need to maintain a positive public image is imperative. Baseball ticket pricing has increased substantially (1) and complex ticket prices could potentially confuse the consumer (7). As a means of determining the best fit in the sport consumer marketplace and how pricing strategies align with peer teams, leagues are examining ticketing price points. This is a simple marketing approach in line with sport marketing professionals (6). Since the advent of the internet price comparison shopping, consumers are able to make buying decisions with a simple click of a button (3, 8). With that said, sport franchises are more conscious than ever of how their ticket prices compare to their competitors’. This research determined, through mainly website analysis, that most of the ticket prices within Double-A and Triple-A baseball affiliate leagues were similar to competition franchises located within their regions. The only exception was found in the Triple-A Pacific Coast League where the South region had substantially lower low-end ticket prices (more similar to that of the Double-A Texas League). As the consumer becomes savvier with online price comparisons, and as economic discretionary spending continues to decline (9), knowing where a team fits within the market offers a greater promotional advantage. Future research may consider examining the impact of how price comparisons can improve sport franchise marketing potential (e.g. illustrating the “value” of minor league entertainment) and measure spectator attitudes toward region price comparisons.

### Applications In Sport

As the present economy is depressed and the future market is unpredictable, discretionary spending on sport entertainment may continue to decline. As such, sport managers within the minor league structure are determining the best approach to continue financial feasibility. This study revealed a common price point for minor league baseball organizations with similar attributes. Most importantly, however, this study revealed that the lowest ticket price in most minor league venues is still relatively affordable. This offers a unique marketing perspective for the increased demand for discretionary spending and sport management organizations should capitalize on this marketing opportunity.

### Acknowledgments

The authors would like to thank graduate research assistant, Lindsey Eidner, and undergraduate research assistant, Nick Garcia, for their invaluable contributions to data collection and analyses of this research endeavor.

### Tables

#### Table 1

Texas League teams organized by region.

Texas League
Double-A Affiliate
Texas Team Non-Texas Team
Corpus Christi Hooks Arkansas Travelers
Frisco Rough Riders Northwest Arkansas Naturals
Midland Rockhounds Springfield Cardinals
San Antonio Missions Tulsa Drillers

#### Table 2

Pacific Coast League teams organized by region.

Pacific Coast League
Triple-A Affiliate
West South Central
Albuquerque Isotopes Nashville Sounds Oklahoma City Redhawks
Fresno Grizzles Memphis Redbirds Colorado Springs Sky Fox
Las Vegas 51’s New Orleans Zephyrs Iowa Cubs
Portland Beavers Round Rock Express Omaha Royals
Salt Lake City Bees
Reno Aces
Sacramento River Cats
Tacoma Rainers

#### Table 3

Descriptive statistics (mean ± standard deviation) for the baseball teams separated by region.

Ticket Prices*
Highest Lowest Fees**
Texas League 13.13 ± 5.40 6.06 ± 0.56 2.13 ± 1.25
  Texas 12.50 ± 4.51 6.13 ± 0.85 2.00 ± 1.08
  Non-Texas 13.75 ± 6.84 6.00 ± 0.00 2.25 ± 1.55
Pacific Coast League 18.13 ± 9.43 6.97 ± 1.19 2.72 ± 1.86
  West 22.63 ± 10.70 7.56 ± 1.05 3.66 ± 2.26
  South 12.00 ± 4.00 5.50 ± 0.58 2.00 ± 0.82
  Central 15.25 ± 6.85 7.25 ± 0.50 1.58 ± 0.15

* Ticket prices are in US dollars
** Fees were team specific, examples included online convenience charges, facility improvement fees, and taxes

#### Table 4

MANOVA source table for the Texas League by region.

Source of Variation df SS MS F
Highest Tickets Between Groups 1 3.13 3.13 0.09
Within Groups 6 201.25 33.54
Total 7 204.38
Lowest Tickets Between Groups 1 0.03 0.03 0.09
Within Groups 6 2.19 0.37
Total 7 2.22
Fees Between Groups 1 0.13 0.13 0.07
Within Groups 6 10.75 1.79
Total 7 10.88

* p < .05

#### Table 5

MANOVA source table for the Pacific Coast League by region.

Source of Variation df SS MS F
Highest Tickets Between Groups 2 345.13 172.56 2.27
Within Groups 13 990.13 76.16
Total 15 1335.25
Lowest Tickets Between Groups 2 11.77 5.88 8.08±
Within Groups 13 9.47 0.73
Total 15 21.23
Fees Between Groups 2 14.33 7.17 2.46
Within Groups 13 37.81 2.91
Total 15 52.14

* p < .05

#### Table 6

MANOVA source table for the Double-A Texas League compared to the Triple-A Pacific Coast League.

Source of Variation df SS MS F
Highest Tickets Between Groups 1 133.33 133.33 1.91
Within Groups 22 1539.63 69.98
Total 23 1672.96
Lowest Tickets Between Groups 1 4.38 4.38 4.11
Within Groups 22 23.45 1.07
Total 23 27.83
Fees Between Groups 1 1.90 1.90 0.66
Within Groups 22 63.02 2.86
Total 23 64.92

* p < .05

### Figures

#### Figure 1

Texas League individual franchise highest and lowest ticket prices.

![figure 1](/files/volume-14/437/figure-1.jpg “figure 1”)

#### Figure 2

Pacific Coast League individual franchise highest and lowest ticket prices.

![figure 2](/files/volume-14/437/figure-1.jpg “figure 2”)

#### Figure 3

Texas League highest ticket price with franchise-specific fees included.

![figure 3](/files/volume-14/437/figure-1.jpg “figure 3”)

#### Figure 4

Pacific Coast League highest ticket price with franchise-specific fees included.

![figure 4](/files/volume-14/437/figure-1.jpg “figure 4”)

### References

1. Alexander, D.L. (2001). Major league baseball: Monopoly pricing and profit-maximizing behavior. Jounal of Sports Economics, 2, 341-355.
2. French, C.W. (2002). Jack Treynor’s ‘Toward a theory of market value of risky assets’. Social Science Research Network. Retrieved April 09, 2010, from <http://papers.ssrn.com/so13/papers.cfm>.
3. Henderson, Dan. (2007, December 13). Online or offline, price comparison tools help consumers shop smart. The Free Library. (2007). Retrieved April 09, 2010, from <http://www.thefreelibrary.com/Online or Offline, Price Comparison Tools Help Consumers Shop Smart-a01073766340>.
4. Jallai, T. (2008). Development of fan loyalty questionnaire for a Double-A minor league baseball affiliate (Master thesis, Texas A&M University-Kingsville, 2008).
5. Lake, C. (2006). Shopping comparison engines market worth £120m-£140m in 2005, says E-consultancy. UK & Global News Distribution. Retrieved April 09, 2010, from <http://www.ukprwire.com/Detailed/Computer_Internet_Shopping_Comparison_Engines_market_worth>.
6. Mullin, B.J., Hardy, S., & Sutton, W.A. (2007). Pricing Strategies. In Human Kinetics (3rd), Sport Marketing (pp. 213-230). Champaign, IL: Human Kinetics.
7. Rascher, D.A., McEvoy, C.D., Magel, M.S., & Brown, M.T. (2007). Variable ticket pricing in major league baseball. Journal of Sport Management, 21, 407-437.
8. Simonds, M. (2009, April 29). Online price comparisons: Easing off your shopping experience. Articlesbase. (2009). Retrieved April 09, 2010, from <http://articlesbase.com/shopping-articles/online-price-comparison-easing-off-your-shopping-experience>.
9. U.S. Bureau of Labor Statistics. (2008). The consumer expenditure survey: Thirty years as a continuous survey.

### Corresponding Author

Liette B. Ocker, Ph.D.
Department of Kinesiology
Texas A&M University – Corpus Christi
6300 Ocean Drive, Unit 5820
Corpus Christi, TX 78412-5820
O (361) 825-2670 F (361) 825-3708
<Liette.Ocker@tamucc.edu>

Using Team Efficacy Surveys to Help Promote Self-and-Team-Efficacy among College Athletes

### Abstract

The purpose of this study was to track and understand attitudinal changes and trends among 3 NCAA Division I intercollegiate teams at the United States Air Force Academy (USAFA). We wanted to see if surveys of team efficacy would help promote self-and-team efficacy with respect to team goals and outcomes. Measures of team efficacy and locus of control were measured throughout the season: preseason, mid-season and postseason. Even though the results varied slightly for each sport, common trends were found with respect to team efficacy and their perceived chances for success and team history. Team goals did not fluctuate much throughout the season. However, results from the survey showed a significant drop in team efficacy for both the baseball and women’s basketball teams from preseason to midseason for both internal locus of control: baseball, t(15) = 3.53, p = .003); women’s basketball, t(15) = 3.67, p = .002. A significant drop in the teams external locus of control was also observed for both baseball, t(15) = 4.43, p < .001 and women’s basketball, t(15) = 2.95, p = .010. However, for the hockey team, there was not a significant drop in internal locus of control, t(15) = 1.23, p = .237 or in external locus of control, t(15) = 1.10, p= .289. As the baseball and women’s basketball teams lost more games both their internal and external locus of control dropped. Accordingly, because the Hockey team did not lose as many games from midseason on their locus of control measures did not experience any drop-off.

**Key Words:** team efficacy, coaching, locus of control

### Introduction

In order to be able to contribute to a team, one must first be confident in one’s own ability to support the team framework. According to Bandura (1) self-efficacy describes the level to which an individual can successfully perform a behavior required to facilitate a specific outcome. Assuming that the individual possesses the skills required to perform the task, self-efficacy is hypothesized to positively influence performance (14). This positive relationship between success and self-efficacy is empirically supported in studies relating to human endurance (25), as well as in the sport of baseball (9). In research targeting task-specific efficacy, supportive evidence suggests that state or task-specific self-efficacy is related to job performance (24) which, in turn, suggests that self-efficacy may also correlate with job performance.

Collective-efficacy has been found to regulate how much effort a group chooses to exert in accomplishing certain tasks, and its persistence in the face of failure (2). Mischel and Northcraft (17) suggested that the cognition of “can we do this task?” is different from the cognition of “can I do this task?” Hodges and Carron (11) and Lichacz and Partington (16), using experimental laboratory tasks, found support for the hypothesis that teams with high collective-efficacy outperformed low-efficacy teams, and that performance failure resulted in lower collective-efficacy on successive performance trials. Prussia and Kinicki (20) also found that collective-efficacy was related to collective goals and performance. Spink (23) found support for a relationship between team cohesion and team-efficacy for elite athletic teams but not for recreational teams. Teams with high collective-efficacy were higher in team cohesion than were teams with low collective-efficacy. Similar studies have indicated that team-efficacy and potency are related positively to performance (10, 13).

Feltz and Lirgg (8) defined team-efficacy as the consensus among players’ perceptions of their personal capabilities to perform within the team. In order to study team-efficacy, Feltz and Lirgg followed one hundred sixty intercollegiate hockey players through the course of a season. They found that team victories increased team-efficacy and team defeats decreased team-efficacy to a greater extent than player efficacy beliefs. They also found a significant decrease in team-efficacy after losing competitions. This opened new doors in the study of team-efficacy because they compared the change in efficacy in times of both success and failure.

#### Attribution Theory

Attribution theory focuses on how people explain their success and failure. According to Weiner, Nierenberg, and Goldstein (26), success and failure are perceived as chiefly caused by ability, effort, the difficulty of the task, and luck. This view, popularized by Weiner, holds that thousands of outside influences for success and failure can be classified into two categories. The first of these categories is stability. Stability is a factor to which one attributes success or failure as fairly permanent or unstable. Factors such as ability, task difficulty, and bias are perceived as relatively stable, whereas other causes, such as luck, effort, and mood are subject to moment-to-moment, periodic fluctuations and are considered unstable.

#### Locus of Control

Locus of control distinguishes two types of individuals: internals, who perceive the likelihood of an event occurring as a product of their own behavior, and externals, who view events as contingent on luck, chance, or other people (22). Causes internal to an individual are ability, effort, and mood. External factors are task difficulty, luck, and bias (26). Team-efficacy includes factors such as team cohesion and the ability of individual players; both internal and stable. Because individuals judge their capabilities partly through social comparisons with the performance of others, it is reasonable to believe that teams will react in the same manner by comparing their collective competencies with their opponents (8,19). Therefore, if a team is comprised of members that ascribe performance success to stable causes, they will expect these outcomes to occur in the future. If team members attribute their success to unstable, external factors, team-efficacy will be much lower. In contrast to these beliefs, Barrick and Mount (5) found no relationship between job performance and stable factors, such as emotion.

In a recent study of NCAA division one baseball players, DeRohan and Nagy (7) found evidence, in support of this theory, suggesting that internal locus of control is more dependent on success and failure then vice versa. Our question centers on a team setting and probes the link between success/failure with locus of control and how they might regulate team-efficacy? The purpose of this study was to track and understand attitudinal changes and trends among 3 NCAA Division I intercollegiate teams at the United States Air Force Academy (USAFA) and see if team efficacy would help promote self-and-team efficacy with respect to team goals and outcomes.

### Methods

#### Study 1: Baseball

##### Participants

Twenty-five male division 1 baseball players and three coaches at the USAFA participated in this study. The player’s ages ranged from eighteen to twenty-four years. Two of the three coaches were in their second year at the Academy. The third coach was in his first year at the Academy. The players and coaches volunteered for this study and did not receive compensation for completing the surveys.

##### Surveys

Preseason, mid-season, and end-of-season surveys were administered to the players and coaches. Each participant took a core of 19-item survey. The surveys measured team-efficacy, attribution theory, locus of control, and demographic information before, during, and after the season of play. The surveys contained restricted-item questions based on six point Likert-type scales ranging from strongly disagree to strongly agree.

##### Procedures

The researchers administered the surveys when the team was all together (e.g., team meetings). The participants took the surveys in the presence of at least one researcher administering each administration of the survey. Upon completion, the researchers collected the surveys from each athlete and coach and placed them in his folder. A mid-season survey was administered the day before the first conference games were played. The same instructions were given as they appeared on the first survey. Upon completion, the administrator collected the surveys from each athlete and coach and placed them in his folder. The same protocols and instructions were followed for the third and final survey that was administered the day after the final conference game was played. Again, upon completion the researchers collected all the surveys and all the data was entered into spss. It is important to note that the coaches never had access to the players’ surveys and the players never had access to the coaches’ surveys throughout the study.

#### Study 2: Basketball

##### Participants

Sixteen female division 1 basketball players from the USAFA basketball team volunteered to participate in this study. One player was eliminated from the study as she did not participate throughout the entire season. Players in this study ranged from 18 to 23 years old. The players and coaches did not receive compensation for completing the surveys.

##### Surveys

The same protocols used for the baseball team were also used for the basketball team. The only difference was the number of surveys administered throughout their season. Whereas the baseball team only had three surveys during their season, the women’s basketball team took a total of six surveys throughout their season: a preseason, four during the season, and one post season survey.

##### Procedures

The same instructions, protocols, and procedures used with the baseball team were used for the women’s basketball team. Upon completion, the researchers collected all surveys for subsequent data compiling and analysis.

#### Study 3: Hockey

##### Participants

Twenty-seven division 1 hockey players and two coaches from the USAFA hockey team participated in this study. Participants ranged in age from 20 to 24 years. All players and coaches agreed to volunteer for this study and did not receive compensation for completing the surveys. All participants were treated according to the American Psychological Association’s ethical guidelines.

##### Surveys

The same core of 19-questions used in the baseball and women’s basketball surveys were also used to survey the hockey team. Like the baseball surveys, the hockey surveys were administered three times during their season: preseason, mid-season, and end-of-the season.

##### Procedure

The same protocols and procedures were used to administer the first two surveys to the hockey team. However, to convenience the hockey players at the end of their season, the researchers gave the postseason surveys to the hockey team captain who agreed to administer it to the team before their final practice leading up to their tournament. Each player took the survey, returned it to the team captain, who then gave them all to the researchers.

##### The Present Study

IRB approval was obtained prior to the start of our investigation. For each team: men’s baseball, women’s basketball, and men’s hockey, we developed two specific hypotheses to address the differences and trends for each team studied at the United States Air Force Academy (USAFA). For example, the baseball team and the women’s basketball team have not been very successful in recent team history. However, based on the trends of wins and losses, the women’s basketball team is on a slightly upward trend in wins whereas the baseball team has maintained a fairly constant trend in wins. The hockey team on the other hand has enjoyed more success in recent history, compiling a significantly higher percentage of wins than the other two teams under study. As a result, we expected to find differing results from these teams based on the team-efficacy levels and their histories of success

### Statistical Analysis

The researchers entered the data generated by the surveys into the Statistical Package for the Social Sciences (SPSS version 14.0) for analysis. Players were then organized according to the last four digits of their social security number. For testing “changes in team-efficacy” and “internal/external locus of control” we created separate constructs. These constructs included internal locus of control, external locus of control, and team-efficacy averages for each survey. In order to establish our efficacy scale, we aggregated the results from eight separate six-point Likert-type scales. Each question targeted team-efficacy individually, but together created the team-efficacy construct. This same process was repeated for the corresponding eight questions in every survey in order to create the efficacy construct for each team. All data were compiled and entered into SPSS for analysis.

#### Baseball

Our first hypothesis was that the margin of victory would be related to team-efficacy. The margin of victory average of each particular two week period was then paired with the respective team-wide team-efficacy score, and graphed as a linear regression (Figure 1). To test our second hypothesis, we ran a dependent samples t-test to measure changes in locus of control between surveys. This gave us 3 dependent samples t-tests for internal locus of control as well as 3 dependent samples t-tests for external locus of control (Figure 2). In order to compare locus of control with margin of victory, we averaged the locus of control scores of all the players to create a team-wide locus of control for each survey.

#### Basketball

Our first hypothesis was the same as for the baseball team: margin of victory would be related to team-efficacy. The margin of victory average of each particular two week period was then paired with the respective team-wide team-efficacy score. To test our second hypothesis we ran a dependent samples t-test to measure changes in players’ internal and external loci throughout the season. In the end, this gave us 15 dependent samples t-tests for internal locus of control as well as 15 dependent samples t-tests for external locus of control. Figure 3 shows the comparisons between internal and external locus of control. In order to compare locus of control with margin of victory, we averaged the locus of control scores of all the players and coaches to create a team-wide locus of control for each survey.

#### Hockey

Our first hypothesis was that winning percentage would be related to team-efficacy over the course of the season. To test this hypothesis, we ran a dependent samples t-test between the efficacy scores from the first survey and from the third survey (Figure 4). As a result the team posted a .472 winning percentage for the season, which was similar to the average for the previous five seasons. As a result, we expected to observe stable team-efficacy scores over the season. In order to test our next hypotheses we ran 3 dependent samples t-tests for internal locus of control, as well as 3 dependent samples t-tests for external locus of control (Figure 5).

### Results

#### Baseball

We found a strong correlation between margin of victory and team-efficacy (r = .907, n = 3). Between the first and second survey, the margin of defeat was 3 while the team-efficacy dropped significantly (t (16) = 5.939, p < .001). Between the second and third survey, however, the margin of defeat was 7 while the team-efficacy dropped only slightly (t (16) = 1.301, p = .212). We found a strong correlation between margin of victory and internal locus of control (r = .785, N = 3). There was also a strong correlation between margin of victory and external locus of control (r = .886, N = 3). Between the first and second survey, the margin of defeat was 3 while internal locus of control dropped significantly (t (16) = 3.526, p = .003) and external locus of control also dropped significantly (t (15) = 4.427, p < .001). Between the second and third survey, the margin of defeat was 7 while internal locus of control dropped only slightly (t (16) = 0.777, p = .448) and external locus of control dropped only slightly as well (t (14) = 1.818, p = .091).

#### Basketball

The correlation between margin of victory and team-efficacy was slightly inversed (r = -0.275, N = 6). Between the first and second surveys, the margin of defeat was 2.4 while the team-efficacy dropped significantly (t (17) = 4.065, p = .001). Between the second and third surveys, the margin of victory was .333 while the team-efficacy dropped just slightly (t (15) = 1.094, p = .291). Between the third and fourth surveys, the margin of defeat was 12.75 while the team-efficacy dropped significantly (t (16) = 3.772, p = .002). Between the fourth and fifth surveys, the margin of defeat was 18.75 while the team-efficacy dropped significantly (t (12) = 3.370, p = .006). Between the fifth and sixth surveys, the margin of defeat was 9.63 while the team-efficacy increased just slightly (t (11) = 0.526, p = .610). The correlation between margin of victory and internal locus of control was weak (r = .296, n = 6), as was the correlation between margin of victory and external locus of control (r = .160, n = 6). Between the first and second surveys, the margin of defeat was 2.4 while internal locus of control dropped significantly (t (17) = 2.50, p = .023) and external locus of control dropped only slightly (t (17) = 0.338, p = .740). Between the second and third surveys, the margin of victory was .333 while internal locus of control dropped significantly (t (15) = 3.674, p = .002) and external locus of control also dropped significantly (t (15) = 2.955, p = .010). Between the third and fourth surveys, the margin of defeat was 12.75 while internal locus of control dropped significantly (t (16) = 3.159, p = .006) and external locus of control also dropped significantly (t (16) = 3.040, p = .008). Between the fourth and fifth surveys, the margin of defeat was 18.75 while internal locus of control dropped only slightly (t (12) = 1.238, p = .239) and external locus of control dropped significantly (t (12) = 2.213, p = .047). Between the fifth and sixth surveys, the margin of defeat was 9.63 while internal locus of control dropped significantly (t (12) = 5.522, p < .001) and external locus of control also dropped significantly (t (12) = 2.243, p = .045).

#### Hockey

From the first to the third survey, the average team-efficacy dropped just slightly from 4.84 to 4.68 (n = 15). Our dependent samples t-test showed us that the team-efficacy was indeed stable (t (14) = 1.451, p = .170). Internal locus of control dropped slightly between the first and second surveys (t (15) = 1.232, p = .237), though it increased slightly between the second and third surveys (t (10) = 1.232, p = .625). Over the entire season it dropped, but not quite enough to attain a significant level (t (13) = 2.042, p = .062). External locus of control increased slightly between the first and second surveys (t (15) = 1.100, p = .289), and dropped significantly between the second and third surveys (t (10) = 2.758, p = .020). Over the entire season it dropped significantly as well (t (13) = 2.842, p = .014).

### Discussion

#### Baseball

Data from the current study showed strong evidence in support of hypothesis one. We found very strong correlations between team-efficacy and margin of victory for each survey distribution. Furthermore, we also found significant differences in team-efficacy between the preseason survey and pre-conference survey. The baseball team had a fairly high level of team-efficacy before the season started. The mean team-efficacy was 4.05 out of 6 possible points. Overall the team as a whole believed they played well together and had the ability to compete successfully at the division one level. However, after winning just 4 games after the first 18, efficacy dropped significantly entering conference play, and remained low for the duration of the season.

The team’s internal locus of control dropped significantly between surveys. Interestingly, external locus of control dropped significantly as well, which we did not predict. Although the results showed no significant difference in changes in internal or external locus of control from the first to the second surveys, both internal and external locus of control dropped significantly when we compared the first and last surveys. We suspect this stems from overall poor performance throughout the season. Perhaps, as the season continued through conference play, the team settled into the reality of both internal factors, such as talent, and external factors, such as their difficult competition.

#### Basketball

We found a slight inverse relationship between the team’s efficacy and margin of victory or defeat for each survey distribution. Although there was not a significant relationship between team-efficacy and margin of victory or defeat, we did find a significant difference in team-efficacy over the season, particularly between the preseason survey and each mid-season survey. The basketball team had a fairly high level of team-efficacy during the preseason. The mean team-efficacy was 4.71 out of 6 possible points. Overall the team as a whole believed they played well together and had the ability to compete successfully at the division one level. However, after winning three of the first eight games, efficacy began to drop. Interestingly, in support of our hypothesis, as the season progressed, team-efficacy levels fluctuated somewhat consistently with the team’s performance, although not necessarily dependent on margin of victory/defeat. It is interesting to note that the freshmen basketball players dropped more significantly in team-efficacy than upper-class players from the preseason survey to the subsequent surveys. The reason behind this might be that the “newer” players were more overconfident before the season began than the upper-class players. It stands to reason that freshmen players entering college are surrounded by athletes better than those they have played with in high school. Because of this, they overestimate the performance potential of the team, and thus report higher levels of team-efficacy before the season begins. The upperclassmen, on the other hand, already have experience in playing at the division one level, and possess a more realistic view of the team’s potential.

As predicted in our second hypothesis, after poor performance between the preseason survey and the mid-season survey, internal locus of control dropped significantly, while external locus of control did not. We predicted a decrease in external locus of control, but not internal locus of control. Interestingly, despite a large margin of defeat between surveys three and four, the data reported significant increases in both internal and external locus of control. However, after taking a closer look at the results of the games, the increase in internal loci makes sense. Although the basketball team experienced two blowout games in a row during conference play, they won a conference game, and then lost the next game by a very close score. Team-efficacy was at its highest after these games, which further explains why they attributed their success to internal reasons.

#### Hockey

As predicted, the results showed that the team had high levels of team-efficacy and showed no significant difference during the season. The results showed a strong relationship between winning percentage and team-efficacy with respect to our linear regression graph. The team lost half of their games between the preseason and mid-season surveys which led to a significant decrease in team-efficacy. Interestingly, the team-efficacy actually increased slightly. Games early in the season are typically non-conference games, where the competition may not be quite as good as the conference teams. However, conference play typically presents a much greater challenge. Therefore, the team felt much more accomplished having won a quarter of the games. Finally, the third survey asked questions that target efficacy over the entire season. Because the questions were based on the whole season, team members apparently considered the season marginally successful.

The hockey team experienced a higher percentage of wins during the season and therefore internal locus of control remained constant throughout the season, which supports our second hypothesis. Despite slight ebbs-and-flows throughout the season, we found no significant difference in internal loci, whereas external loci decreased significantly over the course of the season. Although we predicted that external loci would remain constant, it makes sense that it decreased. Because the team was successful, they did not believe that their success was dependent upon external factors that they could not control.

### Conclusions

The three constructs this present study targeted were team-efficacy, attribution theory, and locus of control within the USAFA baseball, hockey, and women’s basketball teams. For the baseball and women’s basketball teams, the researchers used a margin of victory/defeat construct as a measure of success. Over the past few seasons, the women’s basketball and baseball teams have had very unbalanced win/loss records, which we assumed would not change during the research season. Furthermore, not a single player on the baseball and basketball teams had experienced a winning season, while the hockey team has experienced a higher percentage of wins throughout their seasons. Because of this, we needed something concrete to call “success” that seemed attainable for the teams. For this reason, we assumed that if the teams were losing, but were not continuously getting “blown out” by their opponents, team members would gage those losses as successful. Typically, players take a close loss to a strong opponent much better than a game with a huge point/run spread. In contrast, the hockey team has experienced successful seasons, so we used the team’s win/loss record. Our intent was to run a similar study to that of Feltz and Lirgg (8), and investigate a relationship between team-efficacy and performance. However, specific differences exist between the current study and that of Feltz and Lirgg. First, the current study targeted team efficacy with general questions about team efficacy, whereas Feltz and Lirgg asked statistic-based questions specific to hockey. Second, our results may differ due to the disparity in observations met in each study. Feltz and Lirgg evaluated the team’s level of efficacy after each game during the course of the season. They recorded the team’s efficacy after both wins and losses. Our study only assessed the level of team efficacy during discrete time periods, and we assumed that efficacy carried over between games. While this current study attempted to measure efficacy over the season as did Feltz and Lirgg, constraints in time and resources prevented data collection after each game. The differences in observation may have caused the differing results. Finally, the focus of Feltz and Lirgg’s study was finding that perceived self-efficacy was a strong predictor of performance whereas our study attempted to aggregate self-efficacy into “team efficacy” and performance.

A major limitation of this study was the low sample sizes for each team. Unfortunately, researchers may find this difficult to control because most teams carry fewer than thirty players. To solve this problem, researchers might try to look at multiple teams from the same sport, possibly from different schools (6,15). Future research in this area might find our results to be accurate regardless of our low n-values. As stated in the discussion section above, when a team experiences continual failure, like the baseball and basketball teams in our study, it makes sense that they would continue to regulate external factors for their lack of success, stop attributing their losses to internal factors, and drop in overall team efficacy. This being said, further research should also focus on how wins and losses affect internal locus of control, external locus of control, and team-efficacy. These changes would improve both internal and external validity. The world of sports provides a phenomenal databank of raw information that aids in discovering how people operate in team settings. Taking full advantage of the opportunity to discover as much as possible about the intricate workings of the team atmosphere provides a vital source for improving strategically developed teams in both the corporate and athletic worlds.

### Applications In Sport

Generally speaking, applying the results from these studies to sports seems to reveal that both coaches and players can use these surveys to help regulate their self-efficacy and team efficacy to help monitor their perceived beliefs, motivational levels, and goal attainment throughout the season. As a coach, one could simply use a simple survey to gauge where his or her players are with respect to team efficacy. As a player, one could use the survey to monitor trends in their perceptions of their team’s efficacy and ask themselves why any shifts in perceptions exist. Finally, with respect to locus of control, monitoring team efficacy has the potential to allow players and coaches the opportunity to reflect upon the internal and external influences and how they can change their behaviors to better correspond to their beliefs.

Another application to sport is that winning can indeed affect team-efficacy. Just as we hypothesized, the more a team wins, the higher the team efficacy. Not only does the outcome of winning affect a team’s efficacy, but the margin of victory can also play a significant role. In other words, the more a team wins by (i.e., points, runs, and goals) the greater the team efficacy. Among the coaches however, the drop-off in team efficacy isn’t affected as drastically as the players. Perhaps this has more to do with the “perceived” leadership and how the coaches’ attitudes need to be more even-keeled. In a similar fashion, coaches on winning teams need to maintain a more even-keeled efficacy and not allow themselves to inflate their perception of their team. These applications make good sense in the world of sports psychology with respect to leadership. Ever notice how coaches on championship teams tend to appear to be level-headed and are able to keep their emotions in check?

### Acknowledgments

The authors would like to acknowledge the head coaches and their assistants at the United States Air Force Academy for participating in the studies. Women’s Basketball: Head coach Ardie McInelly , assistant coaches Lisa Robinson, Angie Munger, and Holly Togiai; Men’s Baseball: Head coach Mike Hutcheon, assistant coaches Ryan Thompson, and Scott Marchand; Men’s Ice Hockey: Head Coach Frank Serratore, assistant coaches Mike Corbett, and Andy Berg. We would also like to acknowledge all the student-athletes on these teams who participated in these surveys throughout their respective seasons.

### Tables and Figures

#### Figure 1

Linear regression between margin of victory and team-efficacy for the baseball team.

![figure 1](/files/volume-14/436/figure-1.jpg “figure 1”)

#### Figure 2

Changes in internal and external locus of control for the baseball team between the three administrations of the survey over the course of the season (preseason, mid-season, and postseason).

![figure 2](/files/volume-14/436/figure-2.jpg “figure 2”)

#### Figure 3

Changes in internal and external locus of control for the women’s basketball team between the three administrations of the survey over the course of the season (preseason, mid-season, and postseason).

![figure 3](/files/volume-14/436/figure-3.jpg “figure 3”)

#### Figure 4

Linear regression between winning percentage and team-efficacy for the hockey team.

![figure 4](/files/volume-14/436/figure-4.jpg “figure 4”)

#### Figure 5

Changes in internal and external locus of control for the hockey team between the three administrations of the survey over the course of the season (preseason, mid-season, and postseason).

![figure 5](/files/volume-14/436/figure-5.jpg “figure 5”)

Legend:

1. Preseason team efficacy average
2. Mid-season team efficacy average
3. Postseason team efficacy average

### References

1. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.
2. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, N.J: Apprentice Hall.
3. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman and Company.
4. Bandura, A. (2000). Exercise of human agency through collective-efficacy. Current Directions in Psychological Science, 9, 75-78.
5. Barrick, M. R. & Mount, M. K. (1991). The big five personality dimensions and job performance: A meta-analysis. Personnel Psychology, 44, 1-26.
6. Chase, M. A., Lirgg, C. D., & Feltz, D. L. (1997). Do coach’s efficacy expectations for their teams predict team performance? The Sports Psychologist, 11, 8-23.
7. DeRohan, R. N. & Nagy, C. E. (2005). Team-efficacy, locus of control, and attribution as they relate to a collegiate baseball team. Unpublished manuscript.
8. Feltz, D. L. & Lirgg, C. D. (1998). Perceived team and player efficacy in hockey. Journal of Applied Psychology, 83, 557-564.
9. George, T. R. (1994). Self confidence and baseball performance: A causal examination of self-efficacy theory. Journal of Sport and Exercise Psychology, 10, 381-399.
10. Gully, S. M., Incalcaterra, K. A, Joshi, A., & Beaubien, J. M. (2002). A meta-analysis of team-efficacy, potency, and performance: Interdependence and level of analysis as moderators of observed relationships. Journal of Applied Psychology, 87, 819-832.
11. Hodges, L. & Carron, A. V. (1992). Collective efficacy and group performance. International Journal of Sports Psychology, 23, 48-59.
12. Hysong, S. J. & Quinones, M. A. (1997, April). The relationship between self-efficacy and performance: A meta-analysis. Paper presented at the 12th annual conference of the Society for Industrial & Organizational Psychology, St. Louis, MO.
13. Kellett, J. B., Humphrey, R. H. & Sleeth, R. G. (2000, November). “We’re great” vs. “you’re lazy”: How goal difficulty influences social loafing, collective efficacy and perceived team ability. Paper presented at the annual meeting of the Southern Management Association, Orlando, FL.
14. Kozub, S. A., & McDonnell, J. F. (2000). Exploring the relationship between cohesion and collective efficacy in rugby teams. Journal of Sports Behavior, 23, 120-129.
15. Lau, R.R., & Russell, D. (1980). Attributions in the sports pages: A field test of some current hypotheses about attribution research. Journal of Personality and Social Psychology, 39, 29-38.
16. Lichacz, F. M., & Partington, J. T. (1996). Collective efficacy and true group performance. International Journal of Sport Psychology, 27, 146-158.
17. Mischel, L. J. & Northcraft, G. B. (1997). “I think we can, I think we can …”: The role of efficacy beliefs in group and team effectiveness. Greenwich, CT: JAI Press.
18. Paskevich, D. A., Brawley, L. R., Dorsch, K. R., & Widmeyer, W. N. (1999). Relationship between collective-efficacy and team cohesion: Conceptual and measurement issues. Group Dynamics: Theory, Research, and Practice, 3, 210-222.
19. Pronin, E., Lin, D.Y., & Ross, L. (2002). The bias blind spot: Perceptions of bias in self versus others. Personality and Social Psychology Bulletin, 28, 369-381.
20. Prussia, G. E. & Kinicki, A. J. (1996). A motivational investigation of group effectiveness using social-cognitive theory. Journal of Applied Psychology, 81, 187-198.
21. Roesch, S.C. & Amirkhan, J.H. (1997). Boundary conditions for self-serving attributions: Another look at the sports pages. Journal of Applied Social Psychology, 27, 245-261.
22. Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80, 1-28.
23. Spink, K. S. (1990). Cohesion and collective-efficacy of volleyball teams. Journal of Sport and Exercise Psychology, 12, 3001-311.
24. Stajkovic, A. D. & Luthans, F. (1998). Self-efficacy and work related performance: A meta-analysis. Psychological Bulletin, 124, 240-261.
25. Weinberg, R. S, Gould, D., Yukelson, D., & Jackson, A. (1981). The effect of preexisting and manipulated self-efficacy on a competitive muscular endurance task. Journal of Sport Psychology, 4, 345-354.
26. Weiner, B. (1985). An attribution theory of achievement motivation and emotion. Psychological Review, 92, 548-573.

### Corresponding Author

Andrew D. Katayama, Ph.D.
Department of Behavioral Sciences and Leadership
United States Air Force Academy
2354 Fairchild Dr., Ste 6L101
USAF Academy, CO 80840-6228
<andrew.katayama@usafa.edu>
719-333-1313

Andy Katayama is a Professor of Psychology in the Department of Behavioral Sciences and Leadership at the United States Air Force Academy in Colorado Springs. Andy has also spent six years serving as an officer representative for the intercollegiate varsity baseball team.

Super Bowl Commercial and Game Consumption for the College Demographic

### Abstract

The Super Bowl is the largest annual sporting event in America in terms of single-game television viewership (5). In addition to the game, a tremendous amount of entertainment is intertwined into the Super Bowl telecast via commercials that can cost as much as $3 million for 30 seconds of air time (16). The consumption of the game and commercials is well documented. However, there is little evidence as to how the Super Bowl telecast is consumed by various demographic subgroups. College students, an often overlooked demographic for major sport marketing campaigns, are one group that appear to be an ideal target market for Super Bowl advertising due to their ability for discretionary spending (13) and affinity for popular culture. Therefore, the purpose of this study was to determine the commercial and game consumption patterns for college students during the Super Bowl. A sample of 651 traditional-aged college students at a mid-size Midwestern university was surveyed within 48 hours of Super Bowl XLIV to determine such patterns. Results indicated students watched in large numbers, watched in group settings, identified humor as a primary factor in commercial enjoyment, were interested most in the game itself, identified a different favorite commercial than the USA Today Ad Meter, are strong sport fans, and demonstrated different viewing consumption patterns by gender. It can be concluded from these results that college students resemble the average adult consumer identified in previous research (1,25) in some of their game and commercial consumption patterns (e.g., watching the Super Bowl in large groups and identifying humor as a primary attribute they enjoyed in commercials), but differed in their commercial preferences, their higher level of sport fanship, and their gender differences. Sport marketers can utilize this information to create strategies that appeal to this important demographic.

**Key words:** Super Bowl, commercials, marketing, sport consumption

### Introduction

On February 7th, 2010 the New Orleans Saints defeated the Indianapolis Colts in Super Bowl XLIV. This event, as nearly all Super Bowls before it, came with a tremendous amount of media attention, marketing savvy, and fanfare. In fact, this particular Super Bowl was the second most-watched single-game television program in American history behind

Super Bowl XLV (5).

In addition to the football game, the Super Bowl television broadcast offers a definitive glimpse into the competitive and creative world of sport marketing and advertising. Consumers anxiously anticipate new commercials unveiled during the Super Bowl. These commercials, which cost as much as $3 million per 30 seconds for Super Bowl XLIV (16), have become a cultural phenomenon that create nearly as much buzz as the game itself (1,3,5,16). In essence, the commercials are considered part of the overall Super Bowl entertainment package. McAllister (22), as well as Apostolopoulou, Clark, and Gladden (1), illustrated the integration of Super Bowl commercials into popular culture by examining their content and relative importance. McAllister (2) found that the discourse surrounding the pre, during, and post-Super Bowl advertising led to special status for Super Bowl commercials. Specifically, Super Bowl commercials “often have characteristics more in line with entertainment media messages than stereotypical commercial media messages” (p. 421). Additionally, McAllister explained that Super Bowl commercials are more likely to include celebrities, are much more expensive, and are more thoroughly scrutinized by the public when compared to non-Super Bowl commercials. Blackshaw and Beard (4), echoed these sentiments by noting the uniqueness of Super Bowl advertising is partly due to their entertainment value, their ability to create a “free media” dividend, their high anticipation levels, and their growing ability to engage consumers beyond television (e.g., internet, telecommications, etc.).

Given the elevated status associated with Super Bowl entertainment, Apostolopoulou et al. (1) surveyed 1,101 Super Bowl viewers and NFL database subscribers to determine what elements of the Super Bowl contributed most to their enjoyment. Not surprisingly, the primary contributor to viewer’s enjoyment was the competitiveness of the game itself. The second largest contributor was the specific teams competing, indicating more enjoyment is based on the level of fanship towards a specific team. The third largest contributor was the Super Bowl commercials. The commercials were rated higher than the pre-game show, the celebrity coin toss, the national anthem, the team introductions, the halftime entertainment, and the post-game show. Furthermore, Elliot (9) reported that approximately 4% of the Super Bowl viewing audience watches the Super Bowl only for the commercials. Results from Apostolopoulou et al. (1) and Elliot (9) suggest that beyond the game, commercials have a tremendously powerful influence on Super Bowl viewing patterns.

Although the popularity of Super Bowl commercials is well-documented (1, 4, 16, 22), there is limited evidence to suggest whether a return on investment (ROI) is realized by the companies producing such commercials. Because the Super Bowl is an isolated event that has limited advertising time, in addition to the extreme cost, it is logical to question if ROI is attainable. O’Reilly, Lyberger, McCarthy, Séguin, and Nadeau (25) found a tremendous amount of volatility surrounding the influence of Super Bowl sponsorship. This volatility is caused by the many extraneous factors influencing advertising during this unique event (e.g., presponsorship awareness levels, existing brand associations, increased clutter in the marketplace, etc.). Despite this instability in the marketplace, there has been evidence for increased intent to purchase sponsored products, as well as a willingness on the part of consumers to pay higher prices for goods advertised during the Super Bowl (17, 25, 28). Additionally, an increasing trend for sponsors is to evaluate consumers’ intent to purchase, which ultimately impacts ROI. Blackshaw and Beard (4) noted a clear latency effect on advertised brands during the Super Bowl whereby brand opinion increased 16% and purchase consideration increased 13% in the week following the Super Bowl. Furthermore, the timing of Super Bowl commercials influenced intent to purchase and ROI, where commercials shown closer to the beginning of the game scored higher on nearly every positive advertising measure. Dotterweich and Collins (8) concluded that consumers’ intent to purchase was also impacted by the ratio of value and prestige for any given product. Achieving prestige is often accomplished by repeated brand recognition, and nearly impossible for new companies given the limited television time afforded during the Super Bowl. Therefore, ROI is likely to be greater for companies that are established and have identified a specific target audience versus start-up companies searching for their ideal demographic.

Findings from Dotterweich and Collins (8), as well as O’Reilly et al. (25), are consistent with the results of the USA Today Ad Meter. The Ad Meter is a real-time evaluation of Super Bowl commercials conducted by USA Today whereby participants’ reactions to Super Bowl commercials are measured using a hand-held device. The 2010 Ad Meter gathered information from 250 adult volunteers from San Diego, California and McLean, Virginia. The winning commercial from the 2010 Ad Meter featured famous actress Betty White in a Snickers advertisement. Consistent with findings from Dotterweich and Collins (8), Snickers is an established brand with a certain amount of prestige. Likewise, Anheuser-Busch and its well-known Budweiser commercials have been the Ad Meter champion a record ten times (16). Besides the prestige and brand recognition associated with these Ad Meter winners, there are some other qualities that impact affect toward the advertisement. Kelly and Turley (18) investigated all of the Super Bowl commercials between 1996 and 2002, and used the Ad Meter scores as a dependent variable. Content analysis revealed advertising for goods (i.e., products) was more effective than services, and the use of humor, animals, sports themes, children, and emotional appeal resulted in high levels of affect.

Although the preceding literature helps to contextualize the Super Bowl as a unique advertising opportunity, there is a gap in the literature pertaining to Super Bowl commercial consumption for specific demographic groups. Ad Meter research, as well as general research investigating viewing and consumption patterns have mostly focused their efforts on the general consumer, or on specific commercial content. For example O’Reilly et al. (25) identified respondents from 10 to 94 years of age when assessing their intent to purchase Super Bowl information, and Apostolopoulou et al. (1) investigated adults aged 25 to 44 when examining many forms of Super Bowl entertainment. However, evidence from Zhang, Lam, and Connaughton (30) suggest a need to differentiate demands of various sociodemographic groups during the marketing process. They found the most active sport consumer profile includes individuals that are relatively young (18-25 years of age), single, have low household income, and have a medium entertainment budget. Traditional college students fit this description well.

Given the similarities between the most active sport consumers and traditional college students, the current study attempted to isolate and examine the viewing patterns and perceptions of college students during the most watched sporting event of all time (5). To date, no research has attempted to ascertain the perceptions and consumption patterns of the Super Bowl for this important demographic audience. However, as consumers college students are a powerful force. Oftentimes this demographic, who are currently referred to as millennials (born between 1979 and 1994), are overlooked in marketing outcome research (2). “Considering that college students wield $200 billion in buying power each year, it may be time to set aside any preconceived notions about these coeds and start thinking of them as serious consumers” (13, p. 18). When evaluated individually, it was estimated that the average college student had $287 in discretionary spending per month, which totals $3,444 per year. Additionally, in 2002 over 99% of college students visited the internet a few times per week. Given the nearly exponential growth of the internet, as well as the increase in social and marketing websites, the number of college students who visit the internet regularly continues to rise. It is these technology-savvy college students (2) who are easily reached by the supplemental advertising offered via the Internet, and are often targeted in Super Bowl advertising (e.g., Twitter, Facebook, smart phone applications, etc.).

Within the college student market, as with most forms of marketing research, it is prudent to examine differences in gender consumption patterns. This was a secondary goal of the current research. Developments such as Title IX, female youth sports, and women’s professional leagues have the current generation of sport marketers realizing females are a viable and relevant group of sport consumers with different wants and needs than their male counterparts (24). Females, in general, have demonstrated an affiliation for the feelings of others while fostering communal relationships (27). Work from Chodorow (7) and Gilligan (14) suggests women are more likely to see “morality as emerging from the experience of social connections and value the ethic of responsibility and care.” (6, p. 609). Additionally, female athletes report they most value feelings of belonging, being part of a university community, exercise benefits, and team affiliation (10). These attributes guiding female consumption patterns lend themselves to various marketing strategies, particularly during the Super Bowl when it is common for group viewing to occur, and especially when one considers that nearly half of the viewers of Super Bowl XLIV were female (20). Furthermore, Beasley, Shank, and Ball (3) found that women’s attention levels were higher for Super Bowl commercials than they were for the game itself. According to Zhang et al. (30), “females represent the greatest market potential for professional sports, and identifying their expectations and interests are vital to the future of professional sport organizations” (p. 50).

In contract, males have generally been found to be motivated by an internal self-guided impulse whereby thoughts and behaviors are created by a particular level of self-efficacy or achievement (23). Furthermore, males have been found to be more physically and verbally aggressive (26), more competitive (12), and value autonomy (6). Similarly, male athletes were found to value competition and winning more than the social aspects offered by competitive sport (10), as well as display higher levels of athletic identity (21). From a purely marketing standpoint, males have been found to consume sport more frequently and value specific sport market demands (e.g., win/loss record, team history, close competition, love of the sport, ticket prices, etc.) more than females (30). These gender differences, combined with the unique and powerful college student demographic, may shed light on consumption patterns during the most watched annual sporting event in modern history (5).

### Methods

The purpose of this study was to determine the commercial and game consumption patterns for college students during the Super Bowl.

#### Sample

A sample of 651 traditional college students (mean age = 20.9 years) from a state-funded mid-sized Midwestern university agreed to participate. The original sample consisted of 656 participants, but five surveys were not used due to incomplete or missing answers. A total of 424 males and 227 females were included in the sample. Based on Frankel and Wallen’s (11) sampling methodology, 651 participants was a large enough sample to be representative of the entire University (approximately 20,000 students) at the 99% confidence level.

#### Procedures

Participants located at high-traffic areas on campus (e.g., food courts, busy common areas) completed a nine-item survey designed to determine; 1) The number of Super Bowl commercials watched; 2) Their favorite Super Bowl commercial; 3) The characteristic they enjoyed most about their favorite commercial; 4) Their intent to spend money on any products advertised during the Super Bowl due to the commercial; 5) The number of people they were with when watching the Super Bowl; 6) Their primary interest during the Super Bowl broadcast; 7) Their level of fanship; 8) Gender; and 9) Age. These variables were chosen in an attempt to initiate new lines of inquiry regarding college student consumer behavior s, as well as replicate some of the previous studies on Super Bowl consumption (e.g., 1, 22). Participants were surveyed using a convenience sample of on-campus college students within 48 hours of the completed Super Bowl telecast.

#### Data Analysis

Data analysis was conducted using PASW (version 18). Frequencies and measures of central tendency were used to evaluate all relevant data. Descriptive statistics, Pearson correlations, and multivariate analysis of variance were utilized to determine significance among appropriate categories.

### Results

Research findings are presented in the following four sections: (a) frequencies (b) descriptive statistics, (c) correlations, and (d) MANOVA results.

#### Frequencies

Table 1 represents the frequencies reported by category for each item investigated, and is divided into commercial information and viewing patterns. For the categories identifying favorite commercial and the characteristics of their favorite commercial, the top three answers are provided. Responses for the amount of Super Bowl commercials watched, number of people students were watching with, and level of sport fanship were coded as a number for purposes of further statistical investigation. For example, the amount of Super Bowl commercials watched were coded into four groups where group 1 = none, 2 = a few, 3 = most, 4 = all. These numbers were then used in the following descriptive, correlational, and MANOVA calculations. Age is the only category listed as a mean score.

#### Descriptive Statistics

Table 2 displays descriptive statistics for the number of commercials watched, the number of people present to watch commercials, and the level of fanship. The highest mean score was found for the number of people students were with when they watched the Super Bowl (M = 3.28, SD = .842), indicating the majority of the respondents watched the 2009-2010 Super Bowl within a group setting. Most of the respondents also indicated they watched the majority of all commercials during the Super Bowl (M = 3.14, SD = .697). A high mean score (M = 3.11, SD = .859) for the level of sport fanship indicated the majority of the respondents identified themselves as a ‘big’ or ‘huge’ sports fans. Skewness and kurtorsis values ranged from |.229| to |.553| and |.625|to |.837|, respectively, which were within the criterion of +2.0 indicating establishment of normality among the variables (15).

#### Correlations

Table 3 displays Pearson correlation coefficients for the number of commercials watched, the number of people students were with during the Super Bowl, the level of fanship, gender, and age. A significant positive correlation (r = .21) was found between the number of Super Bowl commercials watched and the level of sport-fanship, indicating individuals who considered themselves big or huge fans were more likely to watch the Super Bowl commercials. A significant negative correlation (r = -.12) was found between the number of commercials watched and gender, indicating male respondents were more likely to watch the Super Bowl commercials. Gender was also negatively and significantly correlated with the level of sport-fanship (r = -.38), indicating male participants identified themselves as stronger sports fans than female respondents. A significant positive correlation between age and the number of commercials watched indicated older students were more likely to watch Super Bowl commercials. Similarly, older students were more strongly identified as a sport fan than younger students (r = .11). As evidenced by a significant positive correlation of .11, the level of sport-fanship was interrelated with the number of people present to watch commercials. This finding indicates that bigger fans were more likely to watch the Super Bowl commercials with people other than students who identified themselves as having a low level of fanship (r = -.09). The highest correlation among the variables was .38, which is lower than the suggested criterion of .85 to establish discriminant validity (19).

### MANOVA Results

To examine gender differences among the factors (the number of commercials watched, the number of people present to watch commercials and the level of sport-fanship – see descriptive statistics across gender in Table 4), a multivariate analysis of variance (MANOVA) was conducted. The results of the MANOVA presented in Table 5 indicate that significant gender differences exist in all factors. The descriptive statistics for the three factors indicated males had higher mean scores in all factors, suggesting males were more likely to be influenced by the three motivational factors listed in the current study. In other words, males were more likely to 1) watch Super Bowl commercials, 2) watch Super Bowl commercials in the presence of others, and 3) display stronger sport fanship. Wilks’ Lambda value of .851 indicated that the three-factor model explained a total variance of approximately 15 percent. However, R squared values at the univariate level indicated that only a small amount of variance (ƞ2 = .014 and .008, respectively) was explained by the number of commercials watched and the number of people present to watch commercials, while a greater amount of variance (ƞ2 = .145) was explained by the level of sport fanship.

### Discussion

The frequency and statistical information presented in Tables 1-5 offers several important insights into the viewing and consumption patterns of college students during Super Bowl XLIV. First, the majority (536 out of 651 participants) indicated they watched most or all of the commercials during the Super Bowl telecast, with males indicating they watched slightly more commercials (M = 3.21) than females (M = 3.04). Given the tremendous popularity of Super Bowl commercials as entertainment (1, 4, 16, 22), it is no surprise college students overwhelmingly engaged in viewing these commercials. However, it is surprising that males watched more commercials than females because males have traditionally been found to be more interested in attributes of the game itself (30), and female attention levels during the Super Bowl were previously found to be higher during commercials (3). Perhaps the Super Bowl is such a mega-special-event (25) that male media desensitization is erased due to the powerful affiliation the commercials seem to have with the game experience. From a practical standpoint, it is important for marketers to acknowledge that male college student demographic is watching the Super Bowl commercials at a greater rate than female college students.

Second, college students were asked to list their favorite Super Bowl commercial and describe the characteristic they enjoyed most about that commercial. With over 70 commercials aired during the Super Bowl XLIV telecast, there were a variety of options. Out of 651 responses, 420 students indicated they had the same favorite commercial. The commercial was from Doritos and it depicted a young boy warning a potential suitor for his mother to “keep your hands off my mama, and keep your hands off my Doritos.” This commercial, although not the Ad Meter winner, was reported by TiVo Inc. to be the commercial that was stopped and played back the most on digital video recorders. Approximately 15% of homes replayed this commercial (5). The results for the Ad Meter ranked this Doritos commercial as the 11th most popular, falling behind another Doritos commercial depicting a dog placing a shock collar on a human to gain access to the man’s Doritos. The second and third rated commercials in the current study (Bud Light ‘house of cans’ and Snickers ‘Betty White football game’) were ranked third and first respectively for the Ad Meter. These results are particularly important for marketers because it suggest a rather large gap (11 spots) between the favorite commercial for college students versus adults who participated in the Ad Meter. This difference validates a marketing strategy that would attempt to isolate particular characteristics of various commercials to be most effective on a specific target audience (e.g., college students).

Although there were different favorite commercials reported for this study and the USA Today Ad Meter, the characteristics that make a commercial enjoyable appear similar. The top three characteristics most enjoyed by college students were humor, creativity, and originality. These characteristics could be used to describe many of the Super Bowl commercials, and reinforces the work of Kelly and Turley (18) who found that humor, sports themes, animals, children, and music increase the likelihood of positive affect towards advertisements. Specifically, the top ranked commercial in this study (i.e., Doritos) used humor and children. The second ranked commercial (i.e., Bud Light house) used humor and music. The third ranked commercial in this study, and first for the Ad Meter (i.e., Betty White football game) used humor and a sports theme. The use of humor is the recurring factor that appears in the top three commercials ranked in this study, and the characteristic chosen by 420 of the 651 participants (64.5%) that made their favorite commercial more enjoyable. Although it is not a secret that humor is an effective advertising tool (also see 29), confirming the importance of this factor within the college student population is a key point for marketers, especially given the overwhelming number of students who reported the importance of humor as the reason they enjoyed their favorite commercial.

Third, this study investigated college students’ intent to spend money on any products advertised during the Super Bowl, commonly referred to as intent to purchase (25). Results confirmed only 11.4% of students indicated they were likely to spend money on any of the products advertised during the Super Bowl. These numbers offer a less than optimistic view for marketers given the intent to purchase is a moderate to strong indicator of ROI (17, 25, 28). Perhaps college students believe the products and services advertised are not important to their lifestyles. Or, perhaps the $3,444 of discretionary spending reported by Gardyn (13) are being applied to products or services that are not advertised during the Super Bowl, but that are important to college students (e.g., laundry, other entertainment, etc.). If this is the case, marketing executives would be wise to understand the general acceptance by college students of Super Bowl advertising, and produce advertising that would peak their interest levels to influence their individual spending behavior, or choose advertising that targets a different demographic altogether.

Fourth, 518 of 651 participants (79.6%) watched the Super Bowl with at least four or more other people, and only 11 students reported watching the Super Bowl by themselves. This finding demonstrates that college students, much like society at large, enjoys the Super Bowl in a communal setting. Furthermore, correlational data demonstrates that males and college students who report themselves as big or huge fans are more likely to watch the Super Bowl with others. The finding that college students watch the Super Bowl in groups is consistent with literature that identifies the Super Bowl as a mega-special-event (25) that focuses on game and surrounding advertisement as a large entertainment package (1). However, the finding that male college students watch the Super Bowl with more people than females is counterintuitive to previous gender literature which suggested females are more attracted to communal relationships and social connections (6, 27). It is possible that the social environment afforded by campus living fosters a more communal context for which males can gather. It is also possible that given the importance of the largest annual sporting event in world, the fans that care about sports most (i.e., males) may find it more important than normal to gather in groups.

Fifth, this study supports the findings by Apostolopoulou et al., (1) which demonstrated that the game itself is the primary point of interest during the Super Bowl. Of the 651 participants, 458 (70.3%) reported they were most interested in the game. This finding is important because it reveals traditional college students are the same as other adults in their primary interest of the game. Marketers can use this interest to incorporate various advertising that might involve the flow of the game (e.g., commercials that air depending on the current score, or commercials that utilize the stars of each team, etc.). Furthermore, the current study found that only 66 of the 651 participants (approximately one percent) tuned into the Super Bowl telecast specifically for the commercials. This finding is noticeably lower than the four percent of the entire Super Bowl viewing audience found by Elliot (9), and implies that college students tune in to specifically watch the commercials at a lower rate than society at large. One must be careful to conclude this result implies college students do not watch commercials. In fact, this study found that most college students do watch and evaluate the majority of Super Bowl commercials, but watching those commercials are not as much of a priority as watching the game itself. Additionally, this study included more males than females, and males have been found to focus more on the aspects of the game (30).

Finally, this study attempted to identify the level of general sport-fanship, and its relationship to viewing patterns. Most students in this study (75.4%) identified themselves as big or huge fans. Males identified themselves as significantly bigger fans than females, and correlational results revealed bigger fans watched more commercials, and watched with more people. These findings allow marketers to begin construction of a blueprint for the average college student sport consumer whereby bigger fans will tend to be males, watch more commercials, and watch with more people. It is not a surprise that males considered themselves bigger fans than females given that males have been found to consume sport more frequently (30), be more competitive (12), value winning more (10), and display higher levels of athletic identity (21). It is also expected that bigger fans would watch more commercials and watch with more people given the large scale entertainment value surrounding the Super Bowl (1, 16). Implications for application are apparent. Marketers should attempt to identify the biggest fans and plan their advertising strategies accordingly. However, one must be mindful that the Super Bowl is a unique event with unusually high consumption, and offers a wide variety of consumers that may not routinely watch sporting events (20). Identifying marketing and advertising strategies for new consumers with varying levels of fanship is a challenging task during any sporting event, but particularly so during the Super Bowl.

### Conclusions

This study evaluated the viewing patterns and perceptions of commercials during Super Bowl XLIV for college students. The results suggest that although college students resemble the average Super Bowl viewer in many ways, they have specific differences that make them an important demographic for marketers to consider. The similarities between college students and the general population include watching the Super Bowl in large numbers, watching the Super Bowl with groups of other people, identifying humor as an enjoyable attribute in advertising, and being interested in the game over other factors (e.g., entertainment). These patterns of behavior lend themselves to specific market segmentation strategies, including advertising that appeals to groups and that contain humor. The differences found between college students and society at large highlight the gap in previous literature that has neglected to isolate this important demographic segment. Specifically, college students displaying a different preference for their favorite commercial (i.e., Doritos house rules vs. Betty White football game) appear to be extremely strong sport fans, and differ greatly by gender. The gender differences are of particular importance because previous literature would lead one to believe that females would be more interested than males in the entertainment portion of the Super Bowl, as well as gathering with large groups of people. This was not the case. The fact that males watched more commercials, and watched in larger groups, implies that male college students are a group of particular importance to marketers. These findings necessitate the need for further research into Super Bowl consumption patterns. Furthermore, given the Super Bowl is the most viewed annual sporting event in the world (5), identifying other sociodemographic consumption patterns is a key for effective marketing strategies.

### Applications in Sports

The results of this study make a strong case for differential game and commercial consumption patterns of college students during the Super Bowl telecast. If sport marketers choose to target the college student demographic in their design of commercials, they would be wise to focus their efforts on strategies which emphasize the importance of the game, using humor as a theme, products that college students would be most likely to use, and concepts that appeal to males. The fact that males are bigger fans, watch both the game and commercials more than females, and gather in groups more so than females, makes the male college student an ideal target market during the Super Bowl. Furthermore, if marketers understand that a majority of college students do not plan to spend money on products or services advertised during the Super Bowl, they may choose to ignore this demographic altogether.

### Tables

Table 1: Total Responses by Category

Variables Category Frequency
Commercial information
Number of Super Bowl commercials watched All 210
Most 326
A Few 115
None 0
Top three favorite Super Bowl commercials 1. Doritos 290
2. Bud Light 72
3. Snickers 45
Top three characters enjoyed most about favorite commercial 1. Funny or humorous 420
2. Creative 24
3. Original 21
Intent to spend money on a product due to a Super Bowl commercial Yes 74
No 489
Not sure 236
Viewing patterns
Number of people students were with when watching the Super Bowl 8 or more 323
4-7 195
1-3 122
0 11
Primary interest during the Super Bowl Game 458
Commercials 66
People students were with 113
Place viewed 14
Level of general sport fanship Huge fan 255
Big fan 236
Somewhat fan 138
Not a fan 22

Table 2: Descriptive Statistics

M SD Skewness Kurtosis
Statistic Statistic Statistic Std. Error Statistic Std. Error
Number of commercials watched 3.14 .697 -.229 .095 -.837 .190
Number of people students were with during the Super Bowl 3.28 .842 -.433 .095 .699 .191
Level of sport-fanship 3.11 .859 -.553 .096 -.625 .191

Table 3: Correlations

(1) (2) (3) (4) (5)
Number of commercials watched (1) 1.0
Number of people present to watch commercials (2) .07 1.0
Level of sport-fanship (3) .21** .11** 1.0
Gender (4) -.12 -.09 -.38 1.0
Age (5) .11** .04 .11** -.04 1.0

* p < .05
** p < .01

Table 4: Descriptive Statistics across Gender Groups

Variables Gender M SD n
Number of commercials watched Female 3.04 .72 227
Male 3.21 .67 424
Number of people present to watch commercials Female 3.18 .91 227
Male 3.33 .80 424
Level of sport-fanship Female 2.67 .81 227
Male 3.35 .78 424

Table 5: Multivariate Analysis of Variance

Source Dependent Variables SS df MS F Sig.a
Gender Number of commercials watcheda 4.270 1 4.270 9.031 .003
Number of people present to watch commercialsb 3.519 1 3.519 4.987 .026
Level of sport-fanshipc 69.142 1 69.142 110.282 .000
Error Number of commercials watched 306.867 649 .473
Number of people present to watch commercials 458.038 649 .706
Level of sport-fanship 406.895 649 .627
Total Number of commercials watched 6754.00 651
Number of people present to wach commercials 7470.00 651
Level of sport-fanship 6775.00 651

Note: Wilks’ Lambda Value = .851; F(3, 647) = 37.635; p < .01
(a) R2 = .014 (Adjusted R2 = .012)
(b) R2 = .008 (Adjusted R2 = .006)
(c) R2 = .145 (Adjusted R2 = .144)

### References

1. Apostolopoulou, A., Clark, J., & Gladden, J. M. (2006). From H-town to Mo-Town: The importance of Super Bowl entertainment. Sport Marketing Quarterly, 15, 223-231.
2. Baker, S. (2004, July 12). Channeling the future. Business Week, 3891, 70-73.
3. Beasley, F. M., Shank, M. D., & Ball, R. W. (1998). Do Super Bowl viewers watch the commercials? Sport Marketing Quarterly, 7(3), 33-40.
4. Blackshaw, P., & Beard, R. (2009). Super Bowl XLIV: Battle for media ROI. Retrieved from Nielsen website: http://enus.nielsen.com/content/dam/nielsen/en_us/documents/pdf/Webinars/SuperBowlWebinar_clientfinal.pdf
5. CBS (2011). Super Bowl sets TV viewership record. Retrieved from http://www.cbsnews.com/stories/2011/02/07/sportsline/main7326154.shtml?tag=mncol;lst;4
6. Chee, K., Pino, N., & Smith, W. (2005). Gender differences in the academic ethic and academic achievement. College Student Journal, 39, 604-619.
7. Chodorow, N. (1978). The reproduction of mothering: Psychoanalysis and the sociology of gender. Berkeley, CA: University of California Press.
8. Dotterweich, D. P., & Collins, K. S. (2005). The practicality of Super Bowl advertising for new products and companies. Journal of Promotion Management, 11(4), 19-31.
9. Elliot, S. (1997, January 24). Advertising. New York Times, p. C5.
10. Flood, S., & Hellstedt J. (1991). Gender differences in motivation for intercollegiate athletic participation. Journal of Sport Behavior, 14, 159-167.
11. Fraenkel, J. R., & Wallen, N. E (2007). How to design and evaluate research in education (6th ed.). New York: McGraw-Hill.
12. Frederick, C. (2000). Competitiveness: Relations with GPA, locus of control, sex, and athletic status. Perceptual and Motor Skills, 90, 413-414.
13. Gardyn, R. (2002). Educated consumers. American Demographics, 24(10), 18-19.
14. Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press.
15. Hair, J. F. Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.). Upper Saddle River, NJ: Prentice-Hall, Inc.
16. Horovitz, B. (2010, February 15). 2010 USA Today Ad Meter tracks Super Bowl XLIV ads. USA Today. Retrieved from http://www.usatoday.com/money/advertising/admeter/2010admeter.htm
17. Jalleh, G., Donovan, R. J., Giles-Corti, B., & Holman, C. D. (2002). Sponsorship: Impact on brand awareness and brand attitudes. Sport Marketing Quarterly, 8, 35-45.
18. Kelley, S. W., & Turley, L. W. (2004). The effect of content on perceived affect of Super Bowl commercials. Journal of Sport Management, 18, 398-420.
19. Kline, R. B. (2005). Principles and practice of structural equation modeling. New York: The Guilford Press.
20. Lapchick, R. (2010, May 6). Super Bowl ads: Time for a change. ESPN.com. Retrieved from http://sports.espn.go.com/espn/commentary/news/story?page=lapchick/100505
21. Lubker, J., & Etzel, E. (2007). College adjustment experiences of first-year students: Disengaged athletes, nonathletes, and current varsity athletes. NASPA Journal, 44, 457- 480.
22. McAllister, M. P. (1999). Super Bowl advertising as commercial celebration. The Communication Review, 3(4), 403-428.
23. Meyers-Levy, J., & Sternthal, B. (1991). Gender differences in the use of message cues and judgments. Journal of Marketing Research, 28, 84-96.
24. Mullin, B., Hardy, S., & Sutton, W. (2007). Sport marketing. (3rd edition) Champaign, IL: Human Kinetics Publishers.
25. O’Reilly, N., Lyberger, M., McCarthy, L., Seguin, B., & Nadeau, J. (2008). Mega- special-event promotions and intent to purchase: A longitudinal analysis of the Super Bowl. Journal of Sport Management, 22, 392-409.
26. Prakash, V. (1992). Sex roles and advertising preferences. Journal of Advertising Research, 32, 43-52.
27. Shani, D., Sandler, D., & Long, M. (1992). Courting women using sports marketing: A content analysis of the US open. International Journal of Advertising, 11, 377-392.
28. Walliser, B. (2003). An international review of sponsorship of events and tax implications: Is there an opportunity for global co-ordination? International Marketing Review, 14, 183-195.
29. Weinberger, M. G., & Gulas, C. S. (1992). The impact of humor in advertising: A review. Journal of Advertising, 21(4), 35-59.
30. Zhang, J. J., Lam, E. T. C., & Connaughton, D. P. (2003). General market demand variables associated with professional sport consumption. International Journal of Sports Marketing & Sponsorship, 5, 33-55.

### Corresponding Author

James E. Johnson, Ed.D.
HP 223A, Ball State University
2000 W. University Ave.
Muncie, IN, 47306
jejohnson1@bsu.edu
765-285-0044

### Author Bio

James Johnson and Donghun Lee are Assistant Professors in the School of Physical Education, Sport, and Exercise Science at Ball State University.

The Selling Mechanism of the Television Rights in Greek Professional Soccer

### Abstract

Television rights in professional soccer was, and perhaps still is, the most important and vital source of revenue for professional soccer clubs in most European countries. Much conversation and legislation was made to discuss, agree upon, and regulate the way the right to broadcast a game is sold to the TV stations and how this income is distributed to the clubs. This study examines the way this selling mechanism works in Greece. The study is carried out with questionnaires, given to at least one member of the higher management of the 34 professional soccer clubs (1st and 2nd division) whose games are on TV. According to the results, club managers think that collective selling is the optimal theoretical model to sell their TV rights, but the way it is implemented is not the optimal one, leading to lower results in income and stadium attendance than the ones anticipated by the managers. Also the Greek Soccer Federation must exploit the TV rights of the games. Moreover the participants believe TV viewers think of the soccer championship as an entity and not as a sum of certain games. Finally they believe TV viewers must pay a subscription to watch the games and that it must not be a free of charge service.

**Key words:** Television rights, Soccer Club Manager, Collecting selling, Individual Selling

### Introduction

While in Greece, studies in the area of professional soccer clubs’ TV rights are very rare, possibly even do not exist, in Europe and especially in the soccer-wise developed countries such as England, Germany, France and Italy. Various studies examined the professional soccer TV rights selling mechanism. Many studies use the USA sports TV market as an example and comparison (since modern sports marketing as we know it was born and developed in the US, and still influences the rest of the world.)

Since soccer became professional and commercialized, the selling of the TV rights was put on the table for discussion. The TV scene in all European countries became free starting from the 1980s until today, and that had much to do with the sharp increase in the professional soccer TV rights value in almost every European country and especially the most developed soccer-wise, such as England, Italy, France and Germany.

Tonazzi (4) points out the differentiality of the soccer market, arguing that while in other business areas a cooperation of the clubs would be characterized as an anticompetitive cartel, in soccer the product can be sold only if the clubs cooperate and offer a joint product. Otherwise, if clubs sell their rights separately, then it is certain that the most popular clubs will gain more income, which would be translated to higher quality players and an unbalanced championship.

Authors argue that if the TV income is lower, then the big clubs most probably will try to separate themselves from the clubs’ league in order to make more money in the free market, selling their TV rights by themselves (13). One probable way to achieve this is the use of digital TV, where they could create tailor-made programs for their fans.

In Europe, England is a leading soccer country and the developments in its televised soccer scene are a case to study, when one wants to define the optimal rights selling mechanism. Poli (14) studies the Italian professional soccer TV rights status in depth.

In Greece, until the early 1990s, EPAE (the governing body of the Greek soccer championship) used to sell the TV rights of the Greek soccer championship collectively to ERT, the public broadcaster. In the early 1990s, private TV stations started bidding to acquire the TV rights of the Greek soccer championship, but it was again the public broadcaster ERT that gained the collective TV rights of the Greek championship. From 1995 to the present, Supersport, a sports channel with subscription fee, has had the TV rights to the majority of the clubs (in 2001 Alpha Digital – a digital platform – acquired the rights for the majority of the clubs, and in the last 3 years ERT has obtained the rights to Olympiakos FC and Xanthi FC). The redistribution system of TV income to the participating clubs is based upon factors like the position of the club in the standings, its market value, its stadium attendance and fans in general, etc.

The purpose of this study is to show that the current TV rights selling model mechanism, used by the Greek professional soccer clubs, is not the optimal one, and revenues and stadium attendance of the clubs could be higher if the way the clubs sell their TV rights were changed. The authors’ hypothesis is that collective selling of TV rights of the Greek professional soccer clubs, based on performance and other criteria (fan base, stadium attendance, etc.), doesn’t maximize the clubs’ revenues or their stadium attendance. The need has been observed for a scientific approach and examination of the TV rights selling mechanism, so that the selling does not only lead to short-term monetary profit, but also to larger, long-term, welfare-wise profits for the parties involved.

### Methods

#### Description of questionnaire – data

For the data collection a questionnaire was used. The questionnaire used in the present research mainly included closed-type questions. The questionnaire was divided in fourteen parts. The first part posed general questions to the participants about the club in which they worked, and the general conditions of Greek professional soccer. Also in this first part, questions about Greek professional soccer’s problems were asked of the club managers. The second part consisted of questions about the ownership of the clubs’ home game TV rights. In the third part the participants were asked about the “product” and the way TV viewers and fans in general view championship and individual games. The fourth part dealt with the supply and demand of the “product”, and the number of games with TV coverage. In the fifth part, the club managers were asked about the cost and profit of the selling mechanism, while in the sixth part the issue was the competitive balance of the championship. The seventh part was a clubs’ talent investment topic and the eighth part questioned the number of club – members in the professional soccer league. In the ninth part, the club managers were asked about the factors influencing the clubs’ decision on choice of selling mechanism (such as stadium attendance, TV households accessibility etc.) The tenth part consisted of questions about the clubs’ TV revenues and their distribution to the clubs, while the eleventh part dealt with regulations and competition policies. Finally in the last part the club managers’ answered social-demographic data questions.

The sample for this study was 65 club managers of the 34 Greek professional soccer clubs who were associated with the clubs during the 2009–2010 season, in the first two divisions (Superleague and Second Division), that are covered by the Greek TV station. The number of clubs of interest was limited, while accordingly limited was the number of the club managers who could answer the questions in this study. Specifically, one to two, or at most three, managers in each club could help in achieving the goal of this study. The number of managers who participated can be easily characterized as quite a large number for this type of study.

#### Statistical Analysis Conducted

Besides the descriptive analysis of single items from the questionnaire, the qualitative variables of the questionnaire were additionally analyzed by utilizing suitable statistical methodology – such as principal components analysis (PCA), and cluster analysis – in order to identify relevant sets of variables and establish a series of factorial (latent) variables that summarize and explain a large proportion of the variability of the observed variables, and logistic regression analysis for attempting to identify the most significant factors for affecting managers’ preferences toward one of the two selling mechanisms.

The data analysis was carried out with the help of the statistical package SPSS v 15.0.

Moreover, in order to see if natural and useful clusters of data existed, the technique of hierarchical cluster analysis was used alternatively. Essentially, starting with each observation being a group by itself, in every step, the observations that have the smaller distance were united, so that the data of a formulated cluster would be part of the elements of the hierarchically next cluster(7,8). This can work not only toward the clustering of observations, but toward the clustering of variables too (7). Since the analysis unit is variable, the distance or similarity measures for all variables’ pairs were calculated. As a distance unit, the Euclidean distance was used and as a method of combination of the observations in clusters the method of “furthest neighbor” was used. According to this method, as a distance between two clusters the one between furthest points was taken (2).

To identify those factors that influence statistically significantly the opinion of Greek managers on the most suitable – according to their own perspective – Greek professional soccer TV rights selling mechanism, a logistic regression model was chosen to fit the data collected (1).

### Results

Analytically, the club managers evaluated the TV rights collective exploitation model to be “very good” (1.5%), “good” (33.8%), “medium” (60%), and “bad” (4.6%). The club members’ answers to the question, “if the TV rights individual exploitation increase the home game stadium attendance” were, “yes” (73.8%), and “no” (26.2%). The club members’ answers to the question, “if the TV rights individual exploitation increase clubs’ income from the TV rights selling” were, “yes” (58.5%) and “no” (41.5%). The club members’ answers to the question, “to whom belong the home games TV rights” were, “to the home team” (3.1%), “to both teams” (6.2%), “to the clubs’ league” (47.7%), and, “to the country’s soccer federation” (43.1%). To the question, “if the TV viewers see the championship as a single product or a sum of independent games” the club managers answered, “as a single product” (80%) and “as a sum of independent games” (20%). To the question, “if the sport product must be treated like a public product and offered free of charge or the viewer to be charged with a subscription fee or other kind of payment” the club managers answered, “like a public product and offered free of charge” (41.5%) and “to be charged with a subscription fee or other kind of payment” (58.5%). To the question, “if the maximization of the clubs’ total profits leads to the maximization of each individual club’s profits” the club managers answered, “a little” (32.3%), “medium” (58.5%), “enough” (4.6%), “much” (1.5%) “very much” (3.1%). To the question, “if the current TV rights selling model has increased, decreased or has not changed the stadium attendance” the club managers answered, “has increased” (36.9%), and “has not changed” (63.1%). Finally, in the instance of the question, “if with the current TV rights selling model of your home games, your revenues, comparing to their real values are higher, equal or lower” the club managers answered, “higher” (24.6%), “equal” (67.7%), and “lower” (7.7%).

In regard to the major problems from which Greek soccer is currently suffering (the means of the sample’s responses on the ten questions range between 2.66 and 3.03), the managers ranked as the most significant problem the lack of suitable training grounds. (see Table 1) The next highest mean value, 3, occured in the response to the question that mentioned the indifference of the State. Lower values showed that the managers considered to be problems the lack of quality of the foreign soccer players and the involvement in the club management of people with no experience in this professional area (2.98), the bad soccer stadiums condition (2.97), the lack of qualitative academies soccer players and the fans’ violence (2.94), with 2.8 the unreliability of the games’ (referees) outcomes (2.8), and the clubs’ bad finances (2.75). The least important problem was regarded by the managers to be the problem of competition with other sports, with a mean value of 2.66, indicating thus the domination of soccer in the Greek sports scene.

#### PCA Analysis

The data that resulted from the items on the questionnaires related to the most significant problems in Greek soccer were given to the club managers of the professional clubs of the Superleague and Second Division in Greece, and then was processed with the main principal components method. The proportion of the variance of each initial variable that the constructed PCA is explained in Table 2. The four principal components comprise 64.3% of the total variability of the ten input variables. For the interpretation of factors, the rotation of factors was conducted. More specifically, the orthogonal transformation process called varimax was used. The objective was to simplify the factor structure and to make the results more meaningful.

The first component showed “the negative attitude of the State and the bodies of professional soccer (Greek Soccer Federation – Referees) toward the ongoing problems of professional soccer (reliability – financial problems)”. The second component showed “the negative correlation that develops between the basic facilities infrastructure of professional soccer and the violence in the Greek stadiums with the foreign players’ quality that professional soccer attracts”. The third component showed “the negative correlation that develops between the professional soccer stadiums’ conditions and the quality of the players coming from the academies into professional soccer”. The fourth component showed “the negative correlation between the competition with other sports and involvement of people with no professional experience in this area in clubs’ management”. (see Figure 1)

#### Cluster Analysis

With the hierarchical analysis in clusters for the problems of Greek soccer, two clusters with the following identities “business type soccer problems” and “soccer problems – involvement of people with no professional experience in this area in clubs’ management” were created. The first cluster mostly dealt with the problems most fans think professional soccer has, and are the main reason of low stadium attendance, low TV viewership, low spending in clubs’ merchandising, etc. Also included was “the competition of the sports”, showing that unhappy fans (mostly those who were not dedicated to the sport) may turn to other sports viewing and attending. The second cluster had more to do with the “structural” problems of professional soccer; that is, the lack of programming and infrastructure in the academies and the training grounds, which leads to the lack of well-trained young and professional players, leading to a low level spectacle on the field. This is a major reason for the fans to turn their backs on their clubs, and on professional soccer in general.

#### Logistic Regression Model

To identify those factors that offer a statistically significantly influence on the opinion of Greek managers on the most suitable – according to their own perspective – Greek professional soccer TV rights selling mechanism, we have chosen to fit a logistic regression model to the data collected. A full description of the predictor variables can be found in Table 4.

A positive evaluation on behalf of the managers of the collective selling mechanism (i.e., “good-very good” category) was designated as predicted group for the dependent variable, while as a reference category the negative category of answers “very bad-medium” was chosen. The maximum likelihood method was used for the adaptation of the final model and the calculation of beta coefficients. In Table 3, the values of the coefficients of independent variables in the logistic model are shown, accompanied by the statistical significance of coefficients, derived by the Wald type test. In the last column, the odds ratios of the model are presented for each of the predictor variables separately.

It follows from an inspection of Table 3that the accessibility of the TV households to the broadcast of the games is a significant factor for the preference of collective selling mechanism, at a 10% level of significance, since those who reported an increase in the accessibility of the households seemed to have lower probability to choose the collective selling mechanism than those reporting the broadcast of games to be left unchanged (beta=-1.623, p-value=0.053<0.1). Indeed, as suggested by the model, the probabilities (odds) of a manager being in a club that had increased the broadcast of its games to TV households to be in favor of the collective selling was decreased by a factor of 0.197, when compared with managers who reported that accessibility was left unchanged. Accordingly, managers whose teams had increased stadium attendance with the utilization of collective selling were less in favor of the current mechanism, when compared with managers whose teams had left its stadium attendance unchanged (beta=-1.537, p-value=0.054<0.1). The most significant factor, however, in predicting the dependent variable in the final model is the club’s revenues. As indicated by the model, the probabilities (odds) of a manager to be in favor of the collective selling model, being in a club that had decreased or left unchanged its revenues with the utilization of the collective selling mechanism was decreased by a factor of 58.997 and 123.304, respectively, when compared with managers who reported that the club’s revenues had increased. (beta=4.077, p-value=0.02

### Discussion

In the study only 34.3% of the club managers considered the current collective selling model to be good or very good. The same clubs’ managers, in the question “whether TV rights individual exploitation increases the home game stadium attendance” answered yes with a rate of 73.8%; and in the question “whether TV rights individual exploitation increased clubs’ income from the TV rights selling” answered yes with 58.5%. This clearly shows a preference of the managers for their clubs to individually exploit their TV rights. As the study showed, managers who were mostly in favor of the individual selling mechanism were those whose teams has been underestimated in revenues compared to their real values. Unexpectedly, these managers believed that the utilization of the current selling mechanism, i.e. collective selling TV rights mechanism, had increased their stadium attendance, and had increased the accessibility of TV households to their games.

Statistical analysis has also shown that the managers who were held a positive stance toward a collective selling model accordingly stated that:

* their TV income with the current collective selling model was the same compared to their real TV rights’ value (73.8%);
* the maximization of total profits of the clubs did not maximize the profits of each club separately (92.9%);
* the less-popular/strong clubs would not get less money with individual selling of their rights (64.3%);
* the financial strengthening of the less-wealthy and -popular clubs, through an even distribution system of TV income, was not among the primary reasons to follow a collective selling model (66.7%);
* the current selling model did not change their team’s stadium attendance (69%);
* the current selling model did not change their club’s financial strength or its ability to acquire talented players (95.2%);
* income distribution based on the clubs’ performances did not change the investment level of the “weak” clubs in talent (66.7%);
* they had considered the possibility of increasing TV ratings of their games in the rights selling procedure (78.6%); and that they
* thought that the accessibility of TV households in games coverage was an important factor in their decision making (85.7%).

Greek soccer experts validated the authors’ hypothesis that the current collective selling model using the performance-based income redistribution system didn’t maximize the clubs’ revenues or stadium attendance.

In past literature, the collective selling mechanism was thought to be the optimal way of exploiting TV rights of the professional soccer championship games in almost all the famous and strong European championships, such as the Premier League in England, Budesliga in Germany, and Division 1 in France. Only in the Italian championship, Lega Calcio, were the TV rights exploited individually, due to the large discrepancies in the predicted and actual revenues of the big and traditional soccer clubs, compared to the small professional clubs (4,9,14,16).

This study shows that the optimal way to exploit Greek professional soccer clubs’ TV rights is via collective selling. That is the model chosen in most of the strongest and most popular professional soccer championships in Europe.

### Conclusions

Based on the findings of the current study, relative studies that were carried out in other European countries, and of course the particularities of the Greek professional soccer market, the authors suggest that the optimal clubs TV rights’ selling mechanism is collective selling through the governing bodies of Greek professional soccer (either the Greek Soccer Federation or the Superleague/EPAE).

The findings of this study clearly showed that the clubs’ managers recognized the need for all the clubs to collectively exploit their TV rights by stating that the games of a championship gain value as part of it, and that TV viewers see the championship as a unity, a product by itself. Mostly it could be concluded by their statement that the Greek soccer federation or the soccer leagues own the clubs’ TV rights and must exploit them. On the other hand, they saw individual selling as a more appropriate model to sell their rights, since in that way they increased their TV income and their stadium attendance. The combination of the two aforementioned contradictory findings, could lead to the conclusion that the club managers think that collective selling is the optimal theoretical model for selling their TV rights, but the way it is currently implemented is not the optimal one, leading to lower results in both income and stadium attendance than those anticipated by the managers. Nevertheless, the need for collective selling was recognized by the managers and by the Greek soccer reality itself, since this model is the model that Greek soccer has chosen to apply for many years, and still does, even now that Greek professional soccer clubs have gained much professional experience by participating in European tournaments and interacting with renowned foreign soccer clubs. The small size of the Greek soccer market and its “hostile” environment to the average fan “client” make this necessity more apparent than ever before.

The current system’s partial failure can be fixed through designing and implementing a “fairer” TV income redistribution model, which will enhance the weaker teams. (As a side note, it is difficult to implement a US-like model that equally distributes the TV income to all the clubs of the league. This is because the whole sport’s theory and concept in USA is totally different than the European one). If weaker teams take more income, then they can afford to acquire better players and create a more competitive squad, leading to a more balanced championship, with more uncertain results. And this uncertainty is the key to league success, through an increase in fans’ interest that is interpreted in higher TV ratings, stadium standings, and spending in soccer products.

### Applications in Sport

This study can be a valuable tool for owners (primarily) and for marketing managers – commercial directors of the Greek professional soccer clubs to compare the Greek TV rights’ selling model efficiency with those used in developed, soccer-wise countries such as England, Germany, France, and Italy. The clubs’ higher management could use this study to evaluate their current selling mechanism, and design and implement a new one that would best fit the Greek soccer market characteristics and have the best possible financial and overall results for all the clubs and the championship.

Specifically, the clubs’ higher management, based on the study, could agree to reform the income redistribution system, so that is not based only on performance related criteria (such as the club’s standing, stadium attendance, fan-base, etc.) and for a more equal distribution of the TV revenues to be applied.

In order to verify the findings, an additional study could be carried out measuring the effect of the current selling model to the TV ratings of the clubs’ televised home games, the TV households’ accessibility in the clubs’ games coverage, etc.

### Figures

#### Figure 1: Dendrogram of the variables of the Greek soccer problems

![Dendrogram of the variables of the Greek soccer problems](http://thesportjournal.org/files/volume-14/434/fig1.jpg)

### Tables

#### Table 1: Ranking of the most significant problems in the Greek soccer by the professional soccer clubs management

Most significant problems in Greek soccer N Minimum Maximum Mean Std. Deviation
Training grounds condition 65 2 4 3.03 0.77
Indifference of the State 65 2 5 3 0.729
Quality of the foreign soccer players 65 2 4 2.98 0.545
Involvement in the club management of people with no experience in this professional area 65 1 5 2.98 0.82
The soccer stadiums conditions 65 2 4 2.97 0.637
Quality of the academies soccer players 65 2 4 2.94 0.726
Fans’ violence 65 2 5 2.94 0.704
Reliability of the game’s outcome (referees) 65 2 5 2.8 0.755
Clubs’ finances 65 2 5 2.75 0.708
Competition with other sports 65 1 4 2.66 0.735

#### Table 2: Results of the PCA model conducted on the items of the questionnaires related to the most significant problems of Greek soccer

Component Initial Eigenvalues Rotation Sums of Squared Loadings
Total % of Variance Cumulative % Total % of Variance Cumulative %
1 2.484 24.836 24.836 2.313 23.131 23.131
2 1.430 14.301 39.137 1.514 15.145 38.276
3 1.328 13.276 52.413 1.351 13.509 51.785
4 1.189 11.893 64.306 1.252 12.521 64.306
5 .927 9.266 73.572
6 .727 7.268 80.840
7 .604 6.045 86.885
8 .531 5.308 92.193
9 .477 4.766 96.958
10 .304 3.042 100.000

Extraction Method: Principal Component Analysis

#### Table 3: Parameter Significance Tests for the logistic regression model for the evaluation of the TV rights income redistribution model (Reference Group: “very bad – medium”)

Parameter Beta Odds Ratio (exp(B))
Intercept n.s.
Collective selling model and stadium attendance (ref.: left unchanged)
Increased positively -1.537* 0.215
Collective selling model and accessibility of the TV households to the broadcast of games(Ref.:left unchanged)
Increased positively -1.623* 0.197
Maximization of the total profits of the clubs and maximization of the profits of each club separately (ref.: very much)
A little n.s.
Moderately n.s.
Enough n.s.
Much n.s.
Collective selling model and the TV ratings of games (Ref.: left unchanged)
Increased positively n.s.
Collective selling model and financial strength/ ability to acquire talented players (Ref.: left unchanged)
Increased positively -1.961**** 0.141
Percentage of Clubs income and TV rights (ref.: 41%-60%)
21-40% n.s.
Collective selling model and club’s revenues (ref.: increased)
Decreased 4,077** 58,997
Left unchanged 4,815*** 123,304
Who must own the home games TV rights (Ref.: Soccer Federation)
Both teams n.s.
Clubs’ League n.s.
-2 Log likelihood 55.961
Nagelkerke R Square 0.488
Cox & Snell R Square 0.355

Dependent Variable: Evaluation of the TV rights income redistribution model of the current collective selling.

* Coefficient is significant at a 10% significance level
** Coefficient is significant at a 5% significance level
*** Coefficient is significant at a 1% significance level
**** Coefficient is significant at a 20% significance level
n.s. Non-significant

#### Table 4: Operationalization of the independent variables used for the logistic regression analysis model

Independent Variables Values
“Who must own the home games TV rights?”
  1. ΗomeClub
  2. Both Clubs
  3. Clubs League
  4. Soccer Federation
  5. Other
“The maximization of the total profits of the clubs leads to the maximization of the profits of each club separately?”
  1. not at all
  2. scarcely
  3. a little
  4. medium
  5. enough
“The current selling model of your TV rights has increased, decreased or left unchanged the TV ratings of your games?”
  1. increased
  2. not changed
  3. decreased
“The current selling model of your TV rights has increased, decreased or left unchanged the accessibility of the TV households to the broadcast of your games?”
  1. increased
  2. not changed
  3. decreased
“The current selling model of your TV rights has increased, decreased or left unchanged your stadium attendance?”
  1. increased
  2. not changed
  3. decreased
“The current selling model of your TV rights has increased, decreased or left unchanged your financial strength and your ability to acquire talented players?”
  1. increased
  2. not changed
  3. decreased
“With the current selling model of your TV rights your revenues, comparing to their real value, are higher, the same or lower?”
  1. higher
  2. same
  3. lower
“What percentage of your club’s income represents the money received from television rights?”
  1. 0-20%
  2. 21-40%
  3. 41-60%
  4. 61-80%
  5. 81-100%

### References

1. El-Hodiri, M. & Quirk, J. (1971). An economic model of a professional sports league. Journal of Political Economy, 79(6), 1302-1319.
2. Fort, R. & Quirk, J. (1995). Cross-subsidization, incentives, and outcomes in professional team sports leagues. Journal of Economic Literature, 33, 1265-1299.
3. Gnardellis, C. (2003). Applied Statistics. Papazisi Publications.
4. Κarlis, D. (2005). Multivariable Statistical Analysis. Stamoulis Publications, Αthens.
5. Kinsella, S. & Smith, H. (1999). Monopoly Structures in Sport, relazione al convegno Sports Broadcasting Rights & EC Competition Law. Paper presented at IBC UK Conferences Limited, London.
6. Mendenhall, W. (1979). Introduction to Probability and Statistics. Fifth Edition. Duxbury Press.
7. Μpechrakis, T. (1999). Multidimensional Data Analysis, Μethods and Applications. Livani Publications.
8. Palomino, F. & Rigotti, L. (2002). The Sport League’s Dilemma: Competitive Balance versus Incentives to Win Tilburg University, Center for Economic Research in its series Discussion Paper with number 2000-109
9. PKF Accountants & Business Advisors in cooperation with Αccountancy Age (2003). Financing Soccer – the New reality. www.ekospor.com/Sports-Finance/04.pdf
10. Poli, E. (2003). The Revolution in the Televised Soccer Market. Italian Media and Telecommunications Authority.Journal of the Modern Italian Studies
11. Siardos, G.K. (1999). Methods of Multivariable Statistical Analysis. Part I. Research of Relations Between Variables. Thessaloniki. Ziti Publications.
12. Tonazzi, A. (2003). Competition policy and the commercialization of sport broadcasting rights: the decision of the Italian Competition Authority. Int. J. of the Economics of Business, 10(1), 17–34.
13. Tsantas, Ν., Μoisiadis, C., Bagiatis Ν. & Chatzipantelis T. (1999). Data Analysis with the help of Statistical Packages. Ziti Publications.

### Corresponding Author

Christos Koutroumanides
Democritus University of Thrace, Greece,
5 Str Dagli, 65403, Kavala, Greece, T 0030-2510-232075
<christoskoutroumanides@yahoo.gr>

### Authors

**Christos Koutroumanides**,
Democritus University of Thrace, Greece

**Athanasios Laios**,
Democritus University of Thrace, Greece

Body Image Disturbances in NCAA Division I and III Female Athletes

### Abstract

The purpose of this study was to examine and compare eating characteristics and body image disturbances in female NCAA Division I and III athletes in the mainstream sports of basketball, softball, track/cross country, volleyball, soccer, tennis, swimming/diving, and ice hockey. Female collegiate athletes (N = 118) from Division I and III universities completed the EAT-26 and MBSRQ. Personal demographics and anthropometric data including height, weight, BMI and Body Fat estimates were also assessed. The study found that 49.2% (Division I) and 40.4% (Division III) of female athletes were in the subclinical eating disorder range. Results assessing body satisfaction, reported that 24.2% of Division I female athletes and 30.7 % of Division III female athletes were either very dissatisfied or mostly dissatisfied with their overall appearance. Results also showed that Division I female athletes were less satisfied with their appearance evaluation (body areas satisfaction, and lower torso). Division III female athletes reported higher levels of bulimic behaviors and weight preoccupation. The results indicate that athletes in refereed female sports are at risk for eating disorders, and that body image risk factors vary between NCAA competition divisions. This research provides sport professionals with a better understanding of risk factors influencing the prevalence of eating disorders between female athletes’ divisional competition levels.

**Key words:** body dissatisfaction, eating disorders, NCAA division, collegiate female athletes, eating disorder risk factors

### Introduction

Eating disorders are among the four leading causes of disease that may lead to disability or death (2). Eating disorders have the highest mortality rate of any mental health illness (41). Approximately nine million Americans suffer from an eating disorder with a lifetime prevalence rate of 0.9% – 4.5% and approximately 10% of college women suffer from a clinical or near clinical eating disorder (19,22).

Body image refers to the self-perception and attitudes an individual holds with respect to his or her body and physical appearance. Body image is a complex synthesis of psychophysical elements that are perpetual, emotional, cognitive, and kinesthetic. Cash and Fleming (10) defined body image as “one’s perceptions and attitudes in relation to one’s own physical characteristics” (p. 455). Body dissatisfaction focuses on body build and is often operationalized as the difference between ideal and current self selected figures (7).

Body dissatisfaction is a significant source of distress for many females. Gender is reported to be a convincing risk factor for disordered eating since females are 10 times more likely to develop an eating disorder compared to males (14). Research shows that the size of the “ideal” woman is far smaller than the size of the average woman (25). “The overwhelming evidence of female gender as a risk factor for the development of an eating disorder highlights the importance of determining the factors that put women at risk, particularly the sociocultural context in which these disorders develop” (31, p. 766).

Risk factors that accompany eating disorders are multi-factorial in nature. Research has revealed that sociocultural, developmental, personality, athletic, trauma, familial, and biological factors are critical identifiable areas that house potential eating disorder risk factors (31). Within these specific areas, body image dissatisfaction and low self-esteem are two situational aspects typically associated with individuals who are at risk for developing an eating disorder. In an early study on body dissatisfaction (5), 23% of the women expressed dissatisfaction with various parts of their body. The particular areas problematic for women were the abdomen, hips, thighs, and overall weight. When the study was replicated in the mid-1980s (11), the percentage of females dissatisfied with their body increased to 38%, with the same general body areas being defined by the participants. These same general body areas were also identified in a more recent study (16) in 56% of women.

Considerable scientific attention has been directed toward the potential role that sport involvement play in an athletes’ development of attitudes and behaviors about disordered eating. Female athletes experience a higher rate of eating disorders than non-athletes (4,24,43). Female athletes have an eating disorder prevalence of 15% to 62% compared to 0.5% to 3% in late adolescent and young adult female non-athletes (21). Researchers (33) assessed disordered eating in female collegiate athletes (N = 204) from three NCAA universities. The responses to the Questionnaire for Eating Disorder Diagnoses (Q-EDD) found 72.5% (n = 148) of the female athletes were asymptomatic, 25.5% (n = 52) symptomatic, and 2.0% (n = 4) eating disorder (29). Compared to recent research (8,39), this research study found a higher percentage of female athletes who were symptomatic. Athlete’s prevalence rate is an important factor, but understanding variables associated with increasing or decreasing risk factors for disordered eating is significant etiological information that should be evaluated (32).

Athletic factors promoting eating disorder development were first identified through research that began in the 1980s, which found particular sports induced higher rates of disordered eating behaviors (1,17). Even though physical activity may develop self-esteem and encourage physical and emotional well-being, there is verification that female athletes are at greater risk for developing disordered eating than their peers who are non-athletes (6). Female athletes encounter the same sociocultural pressures that of non – athletes, however the increased demand of sport – related pressures may independently or dependently increase their risk of eating disordered attitudes and behaviors (40). Coaches, sponsors, and families may all play a role in influencing an athlete’s weight and shape. Negative comments from those that surround and evaluate the athlete may trigger the onset of abnormal eating behaviors leading to an eating disorder (12,28).

The type of sport may also play a role in predisposing an individual to eating disorders based on struggles with body performance satisfaction. Specific sports where performance is judged on body leanness, shape and movement such as ballet, gymnastics, figure skating, diving, and cheerleading have a higher incidence of eating disorders (1,42,47). Shape judged sports such as gymnastics, diving, cheerleading, and dance place more importance on the individual’s body appearance, which may lead to body shape discontent among competitors (47). Researchers also report that 15% to 65% of women in “thin build” sports such as gymnastics or ballet have pathogenic eating patterns known to influence or manipulate the history and development of the eating disorder (27,44). Participation in competitive “thin build” sports in conjunction with personality traits associated with disordered eating could put these individuals at an even greater risk for developing an eating disorder (15, 44). The personality trait of many perfectionist increase disordered eating behaviors for female athletes (20). Researchers (26) compared athletes and non-athletes and reported perfectionism was the only factor that significantly distinguished the groups. In addition, Wilmore (46) reported that athletes high in perfectionism had an increased drive for thinness than athletes low in perfectionism. Refereed sports such as basketball place a stronger emphasis on training and do not rely as much on body appearance; therefore athletes participating in these sports may be less likely to be associated with disordered eating patterns (47).

Most research to date focuses on Division I female athlete’s prevalence rates, while female athletes regardless of NCAA division, experience similar sport specific pressures associated with body image disturbances. Limited research has compared prevalence between NCAA divisions, eating attitudes, and body image disturbances in female athletes. Research has reported that the prevalence of disordered eating, unhealthy dieting, and distorted body image in the athletic population ranges from 12% to 57% (30). Elite female athletes who suffer from eating disorders put themselves at greater risk for serious illnesses and/or death (38). Research has shown that more than one-third of female Division I NCAA athletes report attitudes and symptoms placing them at risk for an eating disorder (2). The National Collegiate Athletic Association study that surveyed student athletes from 11 Division I schools (N = 1,445) reported 1.1% of the female athletes met DSM-IV criteria for bulimia nervosa while 9.2% of female athletes had clinically significant symptoms of bulimia nervosa. This study also reported 0% female athletes met the DSM-IV criteria for anorexia nervosa while 2.85% of the female athletes had clinically significant symptoms of anorexia nervosa (24). Researchers believed the results suggest that Division I female athletes are at significant risk for the progression of eating disorder thoughts and behaviors. The study also stressed the need for future research to examine non-elite Division I, II and III schools since eating disorder risk factors may be higher among lower tier schools. Comparing divisional levels of competition in NCAA athletics could be an important aspect to understanding risk factors involved in the developmental process of an eating disorder.

The purpose of this study was to examine and compare eating characteristics and body image disturbances in female NCAA Division I and III athletes in mainstream sports of basketball, softball, track/cross country, volleyball, soccer, tennis, swimming/diving, and ice hockey. This study also examined female body part dissatisfaction and eating attitudes utilizing the Multidimensional Body Self-Relations Questionnaire (MBSRQ) and Eating Attitudes Test (EAT-26). These findings may assist coaches, strength and conditioning coaches, and athletic trainers in understanding disordered eating and body image disturbances across various female sports in different competition divisions.

### Methods

#### Participants

Participants (N = 118) included Division I (n = 41) and Division III (n = 87) female athletes from National Collegiate Athletic Association (NCAA) member institutes of the following sports: basketball, softball, track/cross country, volleyball, soccer, tennis, swimming/diving, and ice hockey. The convenient sample participants were voluntary, anonymous, and in accordance with university and federal guidelines for human subjects.

#### Instruments

Each athlete completed questionnaires assessing participant demographics and athletic involvement (sport, division). Eating behavior patterns were assessed utilizing the Eating Attitudes Test (EAT-26) and attitudes concerning body image were assessed with the Multidimensional Body-Self Relations Questionnaire (MBSRQ). Anthropometric measurements (height and weight) and body fat measurements were taken on each athlete. (Omron Fat Loss Monitor, Model HBF-306C). The Fat Loss Monitor (Omron Fat Loss Monitor, Model HBF-306C) displays the estimated value of body fat percentage by bioelectrical impedance method and indicates the Body Mass Index (BMI). The bioelectrical impedance, skinfold, and hydrostatic weighing methods have all been shown to be reliable measures of body composition (r = .957 – .987) (23).

##### Eating Attitudes Test (EAT-26)

Eating Attitudes Test (EAT-26) was used to differentiate participants with anorexia nervosa, bulimia nervosa, binge-eating, and those without disordered eating characteristics. It is a 26-item measurement consisting of three subscales: 1) dieting, 2) bulimia and food perception, and 3) oral control. Scoring for this instrument was a Likert scale of six possible answers (always, usually, often, sometimes, rarely, never). Scores ranged from zero to three for each question and a total score greater than 20 indicates excessive body image concern that may identify an eating disorder (Garner et al., 1982; Williamson et al., 1987). EAT-26 has been proven to be a reliable measurement (r = .88) (17). The total score of the EAT-26 and the Drive for Thinness scale of the Eating Disorder Inventory (EDI) have reports of a 90% agreement (37).

##### Multidimensional Body-Self Relations Questionnaire

The Multidimensional Body-Self Relations Questionnaire: The Multidimensional Body-Self Relations Questionnaire (MBSRQ) is a 69 item self-report inventory for the assessment of self-attitudinal aspects of the body image construct. The MBSRQ measures satisfaction and orientation with body appearance, fitness, and health. In addition to seven subscales (Appearance Evaluation and Orientation, Fitness Evaluation and Orientation, Health Evaluation and Orientation, and Illness Orientation), the MBSRQ has three special multi-item subscales: (1) The Body Areas Satisfaction Scale (BASS) approaches body image evaluation as dissatisfaction-satisfaction with body areas and attributes; 2) The Overweight Preoccupation Scale assesses fat anxiety, weight vigilance, dieting, and eating restraint; and 3) The Self-Classified Weight Scale assesses self-appraisals of weight from “very underweight” to “very overweight.” Internal consistency for MBSRQ subscales range from .74 – .91. This questionnaire has been studied and used extensively in the college population. Internal consistency for the subscales of the MBSRQ ranged from .67 to .85 for males and .71 to .86 for females (9).

### Results

#### Descriptive statistics

Participants in the study included 118 female athletes from NCAA Division I (34.7%) and Division III (73.7%) universities. Participants reported their ethnicity as 80.5% White (n =95), 16.1% Black (n =19), .02% Hispanic (n =2), .01% Asian (n =1), and .01% as other (n = 1). The female athletes had a mean age of 19.81 years + 1.29 and a mean body fat percentage of 21.17% + 5.07 (Table 1). There was no significant difference between the divisions in regards to body fat percentage F (1,117) = .727, p = .395.

#### Test for Significance

A multiple analysis of variance (MANOVA) was conducted to determine the effect of NCAA Divisional Status (I or III) on eating characteristics and body image (Table 2). Significant differences were found between Division I and III, Wilks’s Lambda = .664, F(17, 114), p<.0001.

##### Disordered Eating Behaviors

Base frequency scores indicated that 49.2% of Division I female athletes and 40.4% of Division III female athletes scored a 20 or higher on the EAT-26. A follow – up ANOVA reported no significant differences between 20 or higher EAT-26 scores and NCAA Division, F (1, 117) = 1.732, p = .190. A significant difference was found between divisions on the bulimia subscale of the EAT-26, F (1, 117) = 9.107, p = 003. No significant differences were found between division for the EAT-26 dieting subscale, F (1, 117) = .125, p = .724 and oral control subscale F (1, 117) = 2.123, p = .148.

##### Body Disturbance

The results of the MANOVA indicated a significant difference between divisions on the MBSRQ, F(17,114 ) = 3.391, p = .000. The results of the MBSRQ, which assessed body satisfaction, found that 24.2 % of Division I female athletes and 30.7 % of Division III female athletes were either very dissatisfied or mostly dissatisfied with their overall appearance. In addition, a difference was found between Division I and III athletes for appearance evaluation, F (1, 3) = 10.525, p = .001, body areas satisfaction F (1, 3) = 8.36, p = .004, lower torso F (1, 3) = 5.975, p = .016, and overweight preoccupation F (1, 3) = 17.895, p = .000. Division I female athletes were less satisfied with their appearance evaluation, body areas satisfaction, and lower torso than Division III female athletes. Division III female athletes were more weight preoccupied than Division I female athletes.

### Discussion

The main purpose of this study was to examine and compare the eating attitudes and body image satisfaction in female NCAA Division I and III athletes in mainstream sports of basketball, softball, track/cross country, volleyball, soccer, tennis, swimming/diving, and ice hockey. Limited research is available comparing eating disturbances between NCAA divisions so the information acquired may help explain the prevalence of body image disturbances and eating disorder among college female athletes at different levels of competition.

The results of this study indicated that 49.2% (Division I) and 40.4% (Division III) of the female athletes scored 20 or higher on the EAT-26, putting them in a subclinical eating disorder range (18). Comparative research studies using the EAT-26 reported percent subclinical populations of females athletes to be 15.2%, N = 425 (3); 5.8%, N = 190 (13); and 10.2%, N = 59 (36). The current research study did not find a significant difference between subclinical population scores and division, however both Division I and Division III female athletes had a considerably higher subclinical eating disorder female athletic population compared to these previous studies. This finding may be an important implication because the desire to be thin does not always result in clinically diagnosed signs and symptoms of anorexia or bulimia. If left undetected, subclinical eating disorders may result in dysfunctional social interaction, decreased physical performance reduced physical health, and an increase in the propensity for athletic injury.

Between divisions, a significant difference was found on the bulimia subscale of the EAT-26. Division III female athletes struggled more with bulimic behaviors compared to the Division I female athletes. This finding agrees with previous research suggesting that disturbed eating behavior may be higher among lower tiered athletes (35). Bulimic behaviors may be viewed as more destructive to athletic performance so the elite competitive athletes (Division I) may be deterred from participating in such behaviors. Bulimic behaviors may also require a greater level of secrecy, so elite competitive female athletes competing may avoid such behaviors due to increased time commitment, travel requirements, and contact they experience with their coaches and athletic trainers.

It has been reported that female athletes participating in judged sports such as gymnastics, cheerleading, and dance are more prone to eating disorders compared to those who participate in referred sports such as basketball, swimming, and softball (26,34,47). The assessment of body satisfaction through the MBSRQ found that 24.2 % of Division I female athletes and 30.7 % of Division III female athletes were either very dissatisfied or mostly dissatisfied with their overall appearance. We believe that our findings warrant further investigation into the relationship of female athlete’s body dissatisfaction and those participating in referred sports.

A significant difference was also reported on the MBSRQ subscales between Division I and III athletes for appearance evaluation, F (1, 3) = 10.525, p = .001, body areas satisfaction F (1, 3) = 8.36, p = .004, lower torso F (1, 3) = 5.975, p = .016, and overweight preoccupation F (1, 3) = 17.895, p = .000. Division I female athletes were less satisfied with their appearance evaluation, body areas satisfaction, and lower torso than Division III female athletes. Division III female athletes were more weight preoccupied than Division I female athletes. A performance-related drive for thinness through appearance evaluation, body areas satisfaction and lower torso may have a greater impact on female athletes that compete in higher level divisions such as Division I. Being weight preoccupied may not be as closely associated with physical performance measures as compared to general body dissatisfaction.

Even though this was a well-designed study and used a diverse sample of female athletes, it is not without limitations. The participant sample was limited in racial/ethnic minorities, therefore future research should examine female athletic samples with greater racial/ethnic diversity. This research also compared Division I female athletes to Division III female athletes. Increasing the number of institutes and divisions would greatly benefit the findings of this study. Lastly, although a diverse group of female athletic teams was represented in this study, equal number of female athletes from each team was not available due to the sports each institution offered, scholarships, and general participation. For example, ice hockey could only be evaluated at the Division III level. It is possible that the results would have varied if there were equal participant representation. Future research should examine a greater number of institutions at varied divisions to increase participant representations among each sport.

### Conclusion

Our results indicate that refereed female sports are at risk for eating disorders and body image risk factors vary between NCAA competition divisions of female sports. Body dissatisfaction factors that may lead to serious eating disorders will continue to impact the female athletic audience due to added pressures innate to sport performance. Female athletes, regardless of sport, show evidence of risk for developing an eating disorder. Understanding what motivates the developmental process to accelerate in sport may vary depending on level of competition. The educational and scholarly implications of this research project include contributing to the body of literature in the area of body image and eating attitudes of female athletes and providing professionals with a better understanding of the risk factors that influence the prevalence of eating disorders at varied levels of competition.

### Applications in Sport

These findings may assist coaches, strength and conditioning coaches, and athletic trainers in understanding disordered eating and body image disturbances across various female sports in different competition divisions. Professionals that work with female athletes understand the sensitive nature of optimizing performance without compromising overall health. Recognizing and identifying prevention indicators for body image disturbances that lead to disordered eating will assist professionals when dealing with at risk female athletes in varied levels of competition of referred sports. This information will also greatly benefit programs aimed at ceasing the progression of disordered eating

### References

1. Abraham, S. (1996). Characteristics of eating disorders among young ballet dancers. Psychopathology, 29, 223-229.
2. Academy of Eating Disorders. (2008) Statistics and study findings: Burden and prevalence of eating disorders. Retrieved from <http://www.aedweb.org> on January 31, 2008.
3. Beals, K.A., & Manore, M.M. (2002). Disorders of female athlete triad among collegiate athletes. International Journal of Sports Nutrition and Exercise. 12(3), 281 – 293.
4. Berry, T.R., & Howe, B.L. (2000). Risk factors for disordered eating in female university athletes. J. Sport Behavior, 23(3), 207–218.
5. Berscheid, E., Walster, E., & Bohrnstedt, G. (1973) The happy American body: a survey report. Psychology Today, 7(6), 119.
6. Brownell, K.D., Rodin, J., & Wilmore, J.H. (1992). Eating, Body Weight and Performance in Athletes: Disorders of Modern Society. Lea and Febiger, Philadelphia, PA.
7. Candy, C.M., & Fee, V.E. (1998). Underlying dimensions and psychometric properties of the eating behaviors and body image test for preadolescent girls. Journal of Clinical Child Psychology, 27(1), 117-127.
8. Carter, J.E., and Rudd, N.A. (2005). Disordered eating assessment for college student athletes. Women in Sport and Physical Activity Journal, 14, 62–75.
9. Cash, T.F. (2000). The multidimensional body-self relations questionnaire users’ manual. Available from the author at www.body-images.com or through email (TCash@odu.edu).
10. Cash, T.F., & Fleming, E.C. (2002). The impact of body-image experiences: Development of the Body Image Quality of Life Inventory. International Journal of Eating Disorders, 31, 455-460.
11. Cash, T.F., Winstead, B.A., & Janda, L.H. (1985). Your body, yourself: A reader survey. Psychology Today, 20(4), 30-44.
12. Cobb, K.L., Bachrach, L.K., Greendale, G., Marcus, R., Neer, R., Nieves, J. et al. (2003). Disordered eating, menstrual irregularity, and bone mineral density in female runners. Medical Science, Sport, and Exercise, 35, 711-719.
13. Dunn, D., Turner, L.W., & Denny, G. (2007). Nutrition knowledge and attitudes of college athletes. The Sport Journal. 10(4), Retrieved July 5th, 2010, from <http://www.thesportjournal.org/article/nutrition-knowledge-and-attitudes-college-athletes>
14. Fairburn, C.G., & Beglin, S. J. (1990). Studies of the epidemiology of bulimia nervosa. American Journal of Psychiatry, 55, 425-432.
15. Fulkerson, J.A., Keel, P.K., Leon, G.R., & Dorr, T. (1999). Eating disordered behaviors and personality characteristics of high school athletes and nonathletes. International Journal of Eating Disorders, 26, 73-79.
16. Garner, D.M. (1997). The 1997 body image survey results. Psychology Today, 30(1), 30-57.
17. Garner, D.M. & Garfinkel, P.E. (1979). The eating attitudes test: An index of the symptoms of anorexia nervosa. Psychological Medicine, 9, 273-279.
18. Garner, D.M., Olmsted, M.P., Bohr, Y., & Garfinkel, P.E. (1982). The eating attitudes test: Psychometric features and clinical correlations. Psychological Medicine, 12, 871-878.
19. Gurze Books. (2010). Eating disorder statistics. Retrieved July 14th, 2010, from http://www.bulimia.com/client/client_pages/eatingdisorderstats.cfm
20. Hausenblas, H. & Carron, A. (1999). Eating disorder indices and athletes: An integration. Journal of Sport and & Exercise Psychology, 21, 230-258.
21. Hinton, P.S. & Kubas, K.L. (2005). Psychosocial correlates of disordered eating in female collegiate athletes: validation of the ATHLETE questionnaire. Journal of American College Health, 54, 149 – 156.
22. Hudson, J.I., Hiripi, E., Pope, H.G., & Kessler, R.C. (2007). The prevalence and correlates of eating disorders in the national comorbidity survey replication. Biological Psychiatry, 348-358.
23. Jackson, A.S., Pollock, M.L., Graves, J.E., & Mahr, M.T. (1988). Reliability and validity of bioelectrical impedance in determining body composition. Journal of Applied Physiology, 64, 529-534.
24. Johnson, C., Powers, P.S., & Dick, R. (1999). Athletes and eating disorders: The national collegiate athletic association survey. International Journal of Eating Disorders, 26, 79 – 88.
25. Katzmarzyk, P.T., & Davis, C. (2001). Thinness and body shape of Playboy centerfolds from 1978 to 1998. International Journal of Obesity and Related Metabolic Disorders, 25, 590-592.
26. Krane, V., Stiles-Shipley, J.A. Waldron, J., & Michalenok, J. (2001). Relationship among body satisfaction, social physique anxiety, and eating behaviors in female athletes and exercisers. Journal of Sport Behavior. 24(3), 247-265.
27. Macleod, A.D. (1998). Sport psychiatry. Australia New Zealand Journal Psychiatry. 32, 860-866.
28. McLean, J.A., Barr, S.I., & Prior, J. C. (2001). Dietary restraint, exercise and bone health in young women: Are they related? Medical & Science in Sport & Exercise, 33, 1292-1296.
29. Mintz, L.B., O’Halloran, M.S., Mulholland, A.M., & Schneider, P.A. (1997). Questionnaire for eating disorder diagnoses: Reliability and validity of operationalizing DSM-IV criteria into a self-report format. Journal of Counseling Psychology, 44, 1997.
30. Montenegro, S.O. (2006). Disordered eating in athletes. Athletic Therapy Today, 11, 60- 62.
31. Mussell, M.P., Binford, R.B., & Fulkerson, J.A. (2000). Eating disorders: Summary of risk factors, prevention programming, an prevention research. The Counseling Psychologist, 28, 764-796.
32. Petrie, T.A., & Greenleaf, C. (2007). Eating disorders in sport: From theory to research to intervention (pp. 352–378). In G. Tenenbaum & R. Eklund (Eds.), Handbook of Sport Psychology 3rd ed. Wiley & Sons, Inc, Hoboken, NJ.
33. Petrie, T.A., Greenleaf, C., Reel, J., & Carter, J. (2009). Personality and sychological factors as predictors of disordered eating among female collegiate athletes. Eating Disorders: The Journal of Treatment and Prevention, 17, 302 – 321.
34. Powers, P.S. (2000). Athletes and eating disorders. Healthy Weight Journal, 14(4), 59-62.
35. Powers, P.S., & Johnson, C. (1996). Small victories: Prevention of eating disorders among athletes. Eating Disorders: The Journal of Treatment and Prevention. 4, 364-367.
36. Raymond-Barker, P., Petroczi, A., & Quested, E. (2007). Assessment of nutritional knowledge in female athletes susceptible to female athlete triad syndrome. Journal of Occupational Medicine and Toxicology. 2, Retrieved January 27th, 2010, from <http://www.occup-med.com/content/2/1/10>
37. Raciti, M.C. & Norcross, J.C. (1987). The EAT and EDI: Screening, interrelationships and psychometrics. International Journal of Eating Disorders, 6, 579-586.
38. Ryan, J. (1995). Little girls in pretty boxes: The making and breaking of elite gymnasts and figure skaters. New York: Doubleday.
39. Sanford-Martens, T.C., Davidson, M.M., Yakushko, O.F., Martens, M.P., Hinton, P., & Beck, N.C. (2005). Clinical and subclinical eating disorders: An examination of collegiate athletes. Journal of Applied Sport Psychology, 17, 79–86.
40. Smith, A. & Petrie, T. (2008). Reducing the risk of disordered eating among female athletes: A test of alternative interventions. Journal of Applied Sport Psychology, 20, 392 – 407.
41. Sullivan, P.F. (1995)/ Mortality in Anorexia Nervosa. American Journal of Psychiatry, 152(7), 1073 – 1074.
42. Sundgot-Borgen, J. (1994). Risk and trigger factors for the development of eating disorders in female elite athletes. Medical & Science in Sports & Exercise, 14, 59-63.
43. Sundgot-Borgen, J., & Torstveit, M.K. (2004). Prevalence of eating disorders in elite athletes is higher than in the general population. Clinical Journal of Sport Medicine, 14(1), 25–32.
44. Warren, M.P. & Shantha, S. (2000). The female athlete. Baillière’s Clinical Endocrinology and Metabolism, 14, 37-53.
45. Williamson, D., Goreczny, A., & Duchman, E. (1987). Behavioral and psychophysiological assessment of bulimia. Annals of Behavioral Medicine, 9, 8-11.
46. Wilmore, J.H. (1996). Eating disorders in the young athlete. In O. Bar-Or (Ed.), The child and adolescent athlete (pp. 287-303). Oxford, England: Blackwell.
47. Zucker, N.L., Womble, L.G., Williamson, D.A., & Perrin, L.A. (1999). Protective factors for eating disorders in female college athletes. Eating Disorders, 7(3), 207-219.

### Corresponding Author

Kim Kato, Ed.D.
PO Box 13015, SFA Station
Nacogdoches, TX 75962-3015
<kkato@sfasu.edu>
936-468-1610

Dr. Kim Kato is an Assistant Professor in Health Science in the Department of Kinesiology and Health Science at Stephen F. Austin State University in Nacogdoches, Texas.

### Authors

**Kim Kato**, EdD, NSCA-CPT
Stephen F. Austin State University

**Stephanie Jevas**, PhD, ATC, LAT
Stephen F. Austin State University

**Dean Culpepper**, PhD, CC-AASP
Lubbock Christian University