Abstract

The purpose of this study was to examine both the overall and the sports
specific gambling activity among athletes and non-athletes enrolled in
a Southern, regional National Collegiate Athletic Association (NCAA) Division
II university. The findings were contrasted to the results of a 2003 NCAA
Sports Wagering study. The instrument utilized in this particular study
was an adaptation of the survey used in the NCAA 2003 study. Gambling
by athletes at NCAA member schools is a growing concern, and there are
indicators that gambling by college athletes may be more prevalent today
than described in the 2003 study as gambling activity among student-athletes,
male and female, in Division II seems to have increased dramatically from
2003 to 2006.

Specific to this study, respondents from a Southern, regional, NCAA Division
II college, the University of West Georgia, indicated a much higher rate
of gambling as contrasted to the 2003 overall NCAA II findings. Interestingly,
the prevalence of gambling activity among the subjects of this study seemed
to be most prevalent within two sports: women’s basketball and men’s
football. The reported activity among the other nine sports was practically
non-existent. The increase in gambling activity reported by the 2006 student-athletes
as contrasted to the 2003 student-athletes might reflect a change in recreational
lifestyle, ease of access to gambling via the intranet, a rapidly changing
set of sports morays, or an aberration associated with one particular
NCAA II college.

Introduction

Gambling of all types is on the rise in the United States. In 1999, thirty-seven
states and the District of Columbia had lotteries, as compared to thirteen
in 1976 (Claussen & Miller, 2001). In this same year, Nevada hosted
142 bookmaking sites (National Gambling Impact Study Commission, 1999b).
Casino growth has paralleled this expansion of gambling. The approval
rate for using gambling as a way to raise state funds for government programs
and/or education has also dramatically increased. A Gallup Poll conducted
in 1989 indicated that 55% of Americans approved of this type of fund
raising. Ten years later, Goldin (1999) noted that the approval rate grew
to 92% of Americans.
Early acceptance of widespread gambling was evident in The United States
as early as the late 1980s through the passing of the Indian Gaming Regulatory
Act (IGRA). This law gave American Indian Tribes the right to host gambling
activities on reservation grounds as long as the activities were not against
state or national law (Goldin, 1999). Since the passing of this regulatory
act, revenue from gambling has grown from $212 million in 1988 to $6.7
billion in 1999 (NGISC, 1999b). This growth has continued despite disasters
such as “911” and Hurricane Katrina. According to the Mississippi
Gaming Commission (2006), three Gulf Coast casinos alone were able to
generate net revenue of sixty-four million dollars in January of this
year, even after the effects of Hurricane Katrina.

Both private business and governments associated with gambling have responded
to the above phenomena by creating additional opportunities for involvement
with gambling. These include allowing water-based casinos to relocate
to land-based operations; a growth in state lotteries, animal racing,
charitable gambling, video poker machines, sports betting, and internet
gambling (Claussen & Miller, 2001).

Improved technology has created an opportunity for the formation of internet
gambling sites, which lures today’s internet savvy students. Lowry
(1999) reported there were approximately 280 online sites that offer internet
gambling. These online sites generated 1.5 billion dollars revenue in
the year 2000 (Woodruff & Gregory, 2005). Revenue from this type of
gambling will continue to increase as the internet becomes more accessible.
Internet opportunities, along with an increased public acceptance of gambling,
make activities such as betting on an athletic competition more appealing
and much easier (Doocey, 1996; Udovicic, 1998). In fact, sports gambling
has grown to a greater than $100 billion industry (Udovivic, 1998). This
is partially due to the fact that more information (via the Internet)
is available describing sports teams, which allows people to feel more
informed in predicting outcomes.

According to a meta-analysis of gambling habits among university students
by Labrie, Shaffer, LaPlante, & Wechsler (2003), it was reported that
41.9% of students indicated involvement in gambling activity within the
past year, while 23% indicated participation in the activity within the
past week. Additionally, 5.6% of these students met the criteria for pathological
gambling as compared to the rates of 0.2 to 2.1% for the general population
(Labrie, Shaffer, LaPlante, & Wechsler, 2003). This was the result
of a study of college students which utilized findings from the South
Oaks Gambling Screen study (Lesieur & Blume, 1987).

Another study by Engwall, Hunter, & Steinberg (2004) reported similar
findings to the Labrie, et al (2003) meta-analysis. They found that 42%
of college students reported at least one gambling episode in the past
year and 3% of the respondents gambled at least once a week. Labrie, et
al (2003) found that playing the lottery was the most common gambling
activity reported among college students. He found that gambling activity
among college men was significantly greater than college women. Engwall,
et.al (2003) also noted that gambling appears to be related to behavioral
characteristics in college students such as (a) increased television viewing,
(b) computer use for non-academic reasons, (c) spending less time studying,
(d) earning lower grades, (e) participation in intercollegiate athletics,
and (f) binge drinking.

Ironically, alcohol use is a strong predictor of college student gambling
behavior, regardless of gender. Labrie, et al (2003) reported that college
students who had used alcohol within the past year were 2.4 times more
likely to engage in gambling behavior than those who had abstained from
alcohol. It also appears that the variables associated with gambling vary
by gender. For example, among Caucasians, being a male was a strong predictor
of gambling, as contrasted to being a female (Labrie, et.al, 2003). Labrie,
et al (2003) also noted that college female gamblers were more likely
to work for wages, be single, and view community service as less than
very important. Unlike the female gamblers, males who gambled were more
likely to view sports and physical activity as very important.

College students who participated in sports gambling in particular were
more likely to gamble on golf than any other activity according to the
National Collegiate Athletic Association (NCAA) study (Petr, Paskus, &
Dunkle, 2003). Perhaps this is due to the extensive history of gambling
in golf that involves players betting with large sums of money. This has
been documented as far back as 1870 (LeCompte, 2005). The United States
Golf Association (USGA) does not object to gambling that does not interfere
with the game (LeCompte, 2005). This is in contrast to the NCAA policy
that prohibits any type of gambling in the context of athletics.

In a statement to the Senate Commerce Committee, Senator John McCain
noted that college gambling was “reaching epidemic proportions”
(McCain, 2003). Senator McCain made this statement after results from
the National Gambling Impact Study Commission Report (NGISC) indicated
that college students spend more money on gambling activities than alcohol
(NGISC, 1999b).

Gambling on sports by amateur athletes has been added to the list of
behavioral issues addressed by the NCAA. Even though the NCAA prohibits
sports gambling in general, the primary concern has been with participating
athletes betting on games and then shaving points to influence outcomes.
Point shaving has been defined as the deliberate refusal of an athlete
to score in exchange for monetary resources from a book master or “bookie”
(Petr, et al, 2003).

The NCAA utilized the Petr et al. (2003) study to examine the gambling
behaviors of student athletes from all NCAA divisions. The majority of
the activities in which these athletes admitted gambling activity included
playing cards or board games for money, betting on games of personal skill,
purchasing lottery tickets, using slot or electronic poker machines, trading
sports cards, and entering football pools (Petr, et al., 2003).

The overall prevalence of gambling among NCAA student athletes was reported
to be 35 % among males and 10 % among females. Division III athletes were
found to have the greatest prevalence of gambling (Petr, et al, 2003).
In Division I, point shaving was more prevalent among football players
than male basketball players. Just over 1% of football players reported
that they had played poorly in a game in exchange for money, compared
to ½% of the basketball players (Petr, et al., 2003). Golf had
the highest percentage of participants reporting gambling behavior: 8.4
% for females and 48.6% for males.

There appears to be an inverse relationship between knowledge of the
NCAA policy on gambling and the frequency of the behavior. Athletes in
Division III had the highest overall rates of gambling and the least reported
knowledge concerning the NCAA policy on gambling. Only 43.5% of male athletes
in Division III were aware that the NCAA had rules and regulations that
discourage gambling (Petr, et al., 2003), despite the release of the NCAA
publication, Don’t Bet on It. This suggests that this NCAA publication
and the information contained in it may not be disbursed by all member
schools to athletes.

The personality characteristics that produce excellent athletes are also
present in pathological gamblers. These characteristics include feeling
in control of situations and outcomes, a large ego, and optimism (Naughton,
1998). Just as a great athlete is confident in his or her ability to win
competitions, a pathological gambler is confident in accumulating wealth
from gambling. This link alone may account for some of the gambling activity.

There will always be athletes who engage in gambling behaviors despite
being forewarned of the repercussions. The motives have been widely documented;
however, the top stated reasons for gambling by student-athletes have
been reported as “for fun,” “to win money,” and
“for excitement” (Petr, et al., 2003).

Prior to the NCAA sports wagering study conducted in 2003, no data had
been collected specifically looking at the gambling habits of non-Division
I NCAA athletes (Copeland, 2004). The majority of the research on gambling
among athletes has focused on the activities of NCAA Division I men’s
football and basketball players. However, the results of the 2003 NCAA
sports wagering study indicate that additional research on gambling should
be expanded to include athletes in classifications such as Division II
and III. The NCAA study found that 66.5% of Division II athletes, as compared
to 63.4% in Division I, had participated in some form of gambling within
the past year (Petr, et al., 2003). Furthermore, 33.5% of Division II
athletes, as compared to 28.8% of athletes in Division I, admitted participation
in sports wagering within the past year. This is significant, as sports
wagering is prohibited by the NCAA, and results in an athlete losing one
year of eligibility to compete in his or her respective sport if convicted.

While the NCAA (2003) study noted the prevalence of gambling among Division
II athletes, it did not provide data on the specific gambling preferences
of this group nor did it segment the various types of Division II colleges,
such as small-private, large state, or other strata. Therefore, the purpose
of this study was to reexamine the NCAA findings and collect additional,
more current information on the gambling preferences of NCAA Division
II student athletes and non-athletes, with a focus on a NCAA II regional,
rural, state university.

Methods

The subjects selected were all enrolled at the University of West Georgia
(UWG), a NCAA Division II college, during the spring of 2006. This particular
university is a regional school within the University of Georgia System.
The enrollment at the time of the study was approximately 10,800 students.
The subject pool was divided into two groups: non-athlete students and
student-athletes. From each of these two groups a random sample was identified
using alphabetical ordering and then a selection of a predetermined number
of participants based on the total subject pool. The number of athletes
selected was 141 and a 63.1% response rate was obtained thus yielding
fifty-one female and thirty-eight male respondents. The number of non-athlete
students in the initial random sample pool was 220. Eighty-nine or 40.5%
of the subjects agreed to participate, thus yielding a response pool of
fifty-three females and thirty-six males. The predetermined numbers for
the initial subject pools were obtained using the recommendations of Magnani
(1997).

Both student-athletes and non-athlete students completed the survey instrument
in the presence of research assistants. Complete anonymity was guaranteed
and names were not associated with the collected questionnaires. Permission
to conduct the study was granted by the IRB at the University of West
Georgia.

The number of NCAA Division II student athletes participating in the
NCAA Wagering Study conducted by Petr el al (2003) was 1798 females and
2957 males. This data were frequently used for comparison with the findings
of this particular study.

The instrument utilized in this particular study was an adaptation of
the survey used in the NCAA Wagering Study in 2003. The instrument took
between ten and fifteen minutes for the subjects to complete. The survey
instrument identified the prevalence and extent of gambling behaviors
among students within the most recent twelve-month period. Survey items
not completed were labeled by the researchers as “not stated”.
The definition of “not stated” therefore implied the refusal
of the respondent to answer a particular question. In addition to examining
habits, the instrument also identified problem and pathological gambling
behaviors using the South Oaks ten item screening tool as a guide (SOGS).

Results

The findings of this study were tabulated consistent with the format
of the 2003 NCAA wagering study. This allowed for comparisons between
the various categories of respondents. The results were quantified and
examined for observed differences among NCAA II female and male athletes,
UWG female and male athletes, and UWG female and male non-athletes.

As seen in Table 1, among both athletes and non-athletes and females
and males alike, the UWG population in this particular study reported
a higher rate of total gambling activity than the findings in the NCAA
II (2003) female and male athlete population. The UWG population was 18%
to 31% more active in gambling activity in general. However the specific
rate of gambling on college sports at UWG was less than the NCAA II rate
for both females and males.

Table 1

Involvement with Gambling in the Past 12 Months

Note. All NCAA statistics are from Petr et al (2003).
^an= 51 females, 38 males.
^bn= 53 females, 36 males.
^cn= 1798 females, 2957 males.

The findings presented in Table 2 reinforced the sports gambling prevalence
of UWG students and student-athletes. This was particularly true when
wagering on all sports, not just college sports, was considered. Both
UWG females and males were twice as likely to gamble on sports as contrasted
to the total population of NCAA II female and male athletes.

Table 2

Students and Athletes Who Wagered on any Sport by Gender

Note. All NCAA statistics are from Petr et al. (2003).
^a n= 51 females, 38 males.
^b n= 53 females, 36 males.
^c n= 1798 females, 2957 males.

The findings in Table 3 depicted a wide array of gambling activity by
UWG students and student-athletes. This diversity of gambling activity
was evident in the overall NCAA II athlete population as well. Specifically,
non-athlete, female UWG students reported a higher degree of gambling
using card games, whereas male non-athlete UWG students were more likely
to utilize casino table games for gambling purposes. Both female and male
UWG non-athletes were more likely to be involved with craps and dice games
than athletes. The prevalence of gambling activity involving personal
skill was higher among both UWG male and female athletes as contrasted
to non-athletes. Male UWG athletes were twice as likely as any group in
this study to utilize internet gambling options. Utilizing on campus bookies
was three times higher among UWG male athletes as contrasted to all other
groups. The use of off-campus bookies was similar among all groups, except
UWG non-athlete males, who were twice as likely to use an off-campus bookie
compared to the other groups. Female and male UWG students, athletes,
and non-athletes were twice as likely to purchase lottery tickets compared
to the total NCAA Division II group.

Table 3

Students Engaging in Specific Gambling Activities in the Past 12 Months

Note. All NCAA statistics are from Petr et al. (2003).
^a n= 51 females, 38 males.
^b n= 53 females, 36 males.
^c n= 1798 females, 2957 males.

As seen in Table 4, both UWG female and male athletes were nearly twice
as likely to say they had no knowledge of the NCAA gambling rules as contrasted
to the overall NCAA II population responses.

Table 4
Athletes Knowledgeable of the NCAA Rules Concerning Gambling:

Note. All NCAA statistics are from Petr et al. (2003).
^a n= 51 females, 38 males.
^b n= 1798 females, 2957 males.

As seen in Table 5, both female and male athletes at UWG expressed a
similar frequency of problem or pathological characteristics as compared
to those in the NCAA II 2003 study. However, there were a disproportionately
high percentage of non-athlete UWG students whose responses were consistent
with potential problem gambling issues. This group was four times as likely
to indicate potential problem gambling characteristics.

Table 5

Students Who Indicate a Problem or Pathology Concerning Gambling:

Note. All NCAA statistics are from Petr et al. (2003).
^a n= 51 females, 38 males.
^b n= 53 females, 36 males.
^c n= 1798 females, 2957 males.

The authors found that sports gambling athletes from only two sports
among the UWG population displayed significant gambling activity of any
type during the recent twelve months. The sports were women’s basketball
and men’s football. The reported prevalence of gambling activity
among the other nine sports at UWG was not significant.
Discussion

As previously noted, gambling by athletes at NCAA member schools is a
growing concern. The NCAA obviously senses a problem as evidenced by their
focus on the issue. There are indicators that the problem may be larger
than described in the 2003 NCAA study. For example, the fact that 73.2%
of NCAA II athletes in the 2003 NCAA Wagering Study refused to make a
statement about their gambling activity a matter of concern. This could
indicate a fear of being forthright due to concerns about retribution
and conviction.

Also, this study found a much higher rate of gambling among UWG students,
as contrasted to the 2003 overall NCAA Division II population. This could
be more than an aberration associated with one NCAA Division II college.
It could reflect a rapid growth of gambling among college students which
could be related to the widening social acceptance of gambling, the expansion
of internet gambling, or perhaps other issues. However, there is always
the possibility that the limitation due to the smaller number of respondents
among the 2006 UWG population groups, as contrasted to the 2003 NCAA Division
II group, could have skewed the data.

At this point in time however, the UWG population of both athletes and
non-athletes appeared to have a comparatively high rate of gambling involvement.
If one were to assume the rate of involvement among NCAA Division II athletes
has remained constant over the three years since the NCAA study, then
one would have to question whether a regional, rural, relatively large,
state university might have a consistently higher rate of gambling involvement.
This issue alone might merit future study.

Interestingly, the prevalence of gambling activity among UWG athletes
in particular seemed to reside exclusively within two sports, women’s
basketball, and men’s football. The reported activity among the
other nine sports at UWG was practically non-existent. This finding may
be inferable or it might have been the result of a reluctance of athletes
from other sports to express activity among teams. This question also
merits further investigation.

Several other questions associated with gambling among college athletes
merits future study. Is there a link between expressed gambling activity
among student-athletes and graduation rates? Are there athletes from specific
sports that have higher gambling activity rates as indicated in this particular
study? Do non-athlete students actually have a higher gambling activity
rate than the student-athlete population?

Obviously, if gambling becomes an interference with fair
sports competition, the development of the student-athlete, graduation
rates, or the integrity of any aspect of higher education, it deserves
attention. At this point in time, it appears that this determination is
still in question and thus deserves additional research. Additionally,
other universities might consider replicating thus study in order to provide
a basis for comparison and analysis.

References

1. Copeland, J. (2004). Sports wagering survey focuses attention on the high rates of misbehavior in Divisions II, III. The NCAA News. December 6, 2004, Retrieved April 6, 2006 from http://www.ncaa.org/wps/portal/newsdetail
2. Claussen, C.L. & Miller, L.K. (2001). The gambling industry and sports gambling: A stake in the game? Journal of Sport Management, 15, 350-363.
3. Doocey, P. (1996). The case for legal sports betting. International Gaming & Wagering Business, 17 (4), 1, 40-41.
4. Engwall, D., Hunter, R., & Steinberg, M. (2004). Gambling and other risk behaviors on university campuses. Journal of American College Health, 52(6), 245-255.
5. Goldin, N.S. (1999). Casting a new light on tribal casino gaming: Why Congress should curtail the scope of high stakes Indian gaming. Cornell Law Review (Note), 84, 798-849.
6. LaBrie, R., Shaffer, H., LaPlante, D., & Wechsler, H. (2003). Correlates of College Student Gambling in the United States. Journal of American College Health, (52), 53-62.
7. LeCompte, T. (2005). Gambling and golf a match made in heaven. American Heritage, 56 (4), 64.
8. Lesieur, H. & Blume, S. (1987). The South Oaks gambling screen (SOGS): a new instrument for the identification of pathological gamblers. American Journal of Psychiatry, 144(9): 1184-1188.
9. Magnani, R. (1997). Sampling guide. IMPACT Food Security and Nutrition
10. Monitoring Project. Arlington, Va.
11. McCain, J. (2003). Statement of Senator John McCain, Commerce Committee
12. Hearing on Sports Gambling and S.2267, the Amateur Sports Integrity Act.
13. National Gambling Impact Study Commission. (1999b). Final Report. Washington, DC: Author. The Mississippi Gaming Commission. (2006). Monthly Reports. Jackson, MS: Retrieved from www.mstc.state.ms.us/taxareas/misc/gaming/stats/GamingGrossRevenues.pdf April 4, 2006.
14. Naughton, J. (1998). Why athletes are vulnerable to gambling. The Chronicle of Higher Education, 44 (32), A51.
15. Petr, T., Paskus, T.S. & Dunkle, J.B. (2003). NCAA national study on collegiate sports wagering and associated behaviors. National Collegiate Athletic Association, 1-62.
16. Udovicic, A. (1998). Sports and gambling a good mix? I wouldn’t bet on it. Marquette Sports Law Journal, 8 (2), 401-427.
17. Woodruff, C. & Gregory, S. (2005). Profile of Internet Gamblers: Betting on the Future. UNLV Gaming Research and Review Journal. 9 (1), 1-14.

Introduction

A great deal has been written in both academic and popular periodicals
about the value of college athletic programs. While some argue that the
net outcome of college athletic programs is favorable in terms of benefits
to the institution, detractors often view these programs as financially
debilitating to the welfare of the institution (Weeth, 1994). An issue
of controversy for many institutions is the value of the benefits versus
the cost associated with operating intercollegiate athletic programs (Lehnus
and Miller, 1996). The dilemma for administrators is often more pressing
at the small-college level because funding is usually limited and the
programs themselves generally prove to be unprofitable (Helitzer, 1996).
One of the more pressing problems for many small-college athletic programs
is the lack of fan attendance, because attendance can influence support
from alumni and the administration of the school. The present study examines
what factors are key in explaining attendance at small-college sporting
events.

Factors Affecting Attendance

Much research effort has been dedicated to the study of fan attendance
in an attempt to assess fan motivation and other related factors predicting
fan attendance (Wakefield, 1995; Mawson and Coan, 1994; Baade and Tiehen,
1990; Noll, 1974). A number of conceptual and empirical studies have been
directed in the area of sports-fan identity with the team as a future
predictor of attendance (Fisher and Wakefield, 1998; Wann and Schrader,
1997: Zhang, Smith, and Pease, 1996; Pol and Pak, 1994; Yeagin, 1986).
These investigations build upon earlier consumer research in such areas
as group involvement and group identification. Additional streams of sports
marketing research address sports promotion (Helitzer, 1996; Graham, Goldblatt,
and Delpy, 1995; and Wilkenson, 1993) as part of the attendance model.
None of these articles, however, specifically address the promotion of
attendance at small-college athletic programs. Wells et al. (2000) is
one of the few studies that address attendance at small-college sporting
events. They studied small-college football attendance using nine determinants
from DeSchriver’s (1996) model as well as fourteen additional determinants
from a literature review of fan attendance to develop their model. The
significant variables in their analysis were time and season of the game,
winning percentage of the team, promotional effort, prices, whether or
not the school had a sport marketing position, student enrollment, and
the existence of booster clubs.

Data Collection and Analysis

Data were collected at intercollegiate basketball games involving three
small schools in the South and Midwest at approximately the same time
of the season. The questionnaire that was used incorporated much of what
is known or understood to be the salient factors affecting attendance
while including additional factors that were derived from a series of
focus group studies with fans of various sports teams from several small
colleges. It included thirty-nine Likert scale questions (See Exhibit
I for a list of the Likert questions). 492 questionnaires were completed.
Missing data reduced the number of usable questionnaires to 404.

The thirty-nine Likert scale statements (1 = strongly disagree to 5 =
strongly agree) were analyzed using factor analysis to determine their
basic, underlying structure. As described by Hair et al. (1995), eight
of the variables were excluded from the factor analysis because of low
correlations with the other variables. Six factors were extracted, based
on the criterion of having eigenvalues greater than one. The six factors
represented slightly over 55% of the variability in the data. The factor
loadings, after varimax rotation, for the remaining thirty-one variables
on the six factors are shown in Exhibit I; the eight variables not included
in the factor analysis are also described.

Based on the pattern of factor loadings, Factor 1 is labeled “College
Affiliation.” Factor 2 is labeled “Entertainment.” Factor
3 measures the “Affiliation with the Sport.” Factor 4 is “Time
Constraints.” Factor 5 is a measure “Team Familiarity.”
Finally, Factor 6 is “Lack of Awareness.”

The purpose of the factor analysis was to use the results in a regression
model to explain attendance. As described by Hair et al. (1995) surrogate
variables, summated scales, or factor scores might be used for this purpose.
For this study, factor scores were used. The independent variables in
the model were therefore the six factors described above, using the corresponding
factor scores, and a number of dummy variables: GENDER (the gender of
the respondent; 0 = male, 1 = female), MARITAL (marital status; 0 = single,
1 = married), and CHILDREN (whether the respondent has children; 0 = no,
1 = yes). Finally, the eight Likert scale variables that were eliminated
from the factor analysis were included.

The dependent variable, which is the number of home games attended (GAMES),
is a series of discrete values from 1 to 5 (1 = first home game, 2 = 2
home games, 3 = 3 or 4 home games, 4 = 5 to 7 home games, inclusive, and
5 = 8 or more home games). The distribution of GAMES is shown below.

An appropriate regression procedure when the dependent variable is ordinally
scaled is ordered probit. Therefore, in order to examine the effects of
the independent variables on attendance, Minitab’s? ordered probit
procedure was used with GAMES as the dependent variable and with the factor
scores for the six factors and the other independent variables as described
above. The results were that only the six factors were statistically significant.
Therefore, another ordered probit model was created using only the six
factors; the results are shown below. The model is statistically significant
based on the G statistic, which follows a ?2 distribution with the degrees
of freedom equal to the number of independent variables (Hosmer and Lemeshow,
1989).

Log-likelihood = -495.180
Test that all slopes are zero: G = 239.220, DF = 6, P-Value = 0.000

Factors 1 through 6 are all significant using a 5% alpha value. Because
of the way Minitab? calculates the coefficients in ordered probit analysis,
the reported negative coefficients indicate that an increase in the independent
variable tends to be associated with a greater attendance. The pattern
of coefficients is as one would expect.

In linear regression, the estimated coefficients can be interpreted as
marginal effects. In ordered probit, the marginal effects must be calculated
using the coefficients, and are reported as probabilities. The marginal
effects were calculated and resulted in importance ranking of the factors
that were the same as the absolute value of each factor’s coefficient.
Therefore, the importance ranking of the six factors, from most to least,
is Factor 5 (Team Familiarity), Factor 1 (College Affiliation), Factor
3 (Affiliation with the Sport), Factor 6 (Lack of Awareness), Factor 4
(Time Constraints), and Factor 2 (Entertainment).

Discussion

Factor 1: College Affiliation

Research within the social science discipline indicates that peer group
affiliation creates a sense of belonging and identity (Parsons, 1993).
While secondary group affiliation plays a smaller role in the individual’s
identity and affiliation in terms of group dynamics, individual membership
and a sense of belonging are important to the formation of organizational
cultures. Larger organizational groupings do tend to play a major role
in the development of the type of organizational culture thought to exist
on many college campuses. Secondary group membership has been closely
linked with both organizational culture and the development of esprit
de corps within the organizational structure (Hunt, Wood, and Chonko,
1989; Tajfel, 1981). As Wakefield (1995) has indicated, attending a sporting
event is a highly social event, and thus the effects of reference group
acceptance may be considered a determining factor in patronage intentions.
Murrell and Dietz (1992) have also indicated that fans who maintain a
strong identity with a university as their relevant institution, will
manifest that identification in greater support for the school’s
sports teams. In the present study, Factor 1 (College Affiliation) was
the second most important factor influencing attendance, suggesting that
individual association with a school has a powerful effect on attendance
at school sponsored sporting events.

Factor 2: Entertainment

The Entertainment factor was the least important influence on attendance.
Entertainment included special events, prizes, and sales promotions designed
to increase excitement and attendance. Research on the actual effect of
promotional activities on sport attendance is varied even though promotion
of sporting events is considered an essential element of success for any
sport franchise. Promotional activities, however, have been demonstrated
to produce mixed results. While some teams experience increased sport
attendance figures throughout the season as a result of the team’s
promotional activities, other teams have discovered that much of what
constitutes an “increase” is in fact temporal. The net effect
of season long stimulation for the purpose of increasing patronage is
that that a marketing barrage only affects those people who attend solely
for the purpose of receiving the sort of novelty item being offered at
a “special event” (Pitts and Stotlar, 1996). Hence, there
is a fine line between drawing attention to the team (or to the sporting
event) and interrupting the normal attendance schedule through promotional
activities. Promotions can either be considered an effective method of
demonstrating appreciation to the everyday sport consumer, or they can
mask serious deficiencies in actual fan support.

Factor 3: Affiliation with the Sport

One of the more obvious reasons why individuals would choose to attend
a sporting event is because they enjoy the sport itself. People who are
fans of a sport have developed a fondness for the intricacies of the game
and are more likely to choose to further their own participation in the
sport by becoming fans. Krohn and Clarke (1998) indicate that people who
attend sporting events can be characterized either as spectators or fans.
While spectators fulfill their enjoyment by casually viewing the sport
and not getting caught up in the logistics of the event, most true fans
attend sporting events because of some deep involvement in what the authors
describe as “the almost religious rituals” one sometimes associates
with the sporting event itself. While there are many ways of developing
an interest in a sport, one of the principal methods of developing deep
knowledge of a sport is through participation, either as a player or as
a spectator.

Factor 4: Time Constraints

This was the fifth most important factor in the model. In order for sporting
events to become attractive enough so that they become an integral part
of the fan’s schedule, the scheduling of the events should coincide
with the lifestyle and schedule of the primary attendees. The timing of
a sporting event is important in that if it is not conducive to the time
constraints and scheduling conflicts of the primary fan base, then the
event will not be well attended. However, it could be argued that time
conflicts are an excuse for not attending. A true fan would find how to
attend in spite of conflicts.

Factor 5: Team Familiarity

This was the most important influence on attendance in the model. Fan
identification with players of a particular sports team is an area in
which personal commitment and emotional involvement by the fan often occurs.
In rare cases, fans have so closely identified themselves with an organization’s
players that they begin to define themselves in terms of the attributes
of those players (Mael and Ashforth, 1992). Wann and Branscombe (1993)
have found that high fan identification with a team and its players relates
to additional involvement with the team, which in turn relates to greater
attendance at home games. In general, sport as a whole is thought to differ
significantly from other forms of entertainment because sports tend to
evoke a higher level of emotional attachment and identification from its
fans (Sutton, et al., 1997). As Lever (1983) indicates, sport not only
promotes communication among people, it tends to involve diverse groups
of people by providing common symbols and a collective sense of solidarity
for both the players and the sports organization.

Factor 6: Lack of Awareness

College athletic departments share the common need of promoting their
own product, in this case, the sporting event itself. Ironically, advertising
the event and promoting the general awareness of the scheduled time of
play and the opponent during the contest is not listed as the top perceived
priority of athletic department marketing personnel. Instead, college
athletic department marketing personnel list the job of selling corporate
sponsorships as their top priority. The second most important job responsibility
(as identified by 52% of athletic directors) is the planning and implementation
of individual game promotions, followed closely (at 48%) by planning and
directing season-ticket campaigns (Lehnus and Miller, 1996).

Respondents mentioned the general lack of awareness and knowledge of
the time of the sporting event and lack of awareness and knowledge about
the identity of the opponent as possible factors for why fans failed to
show for the game but, as with time conflicts, this may simply be an excuse.
Real fans would learn about the schedule.

Conclusions and Strategy Recommendations

An interesting outcome of this study is the relatively low importance
of win/loss records in explaining attendance. Only one of the Likert questions
(Q37: “I would not attend <SCHOOL> basketball games if the
team were not winning) was used in the factor analysis, and it loaded
(loading = 0.398) on the Entertainment factor. The three other questions
concerning the records of the teams (Q36: “One of the main reasons
I attend <SCHOOL> basketball games now is because of the team’s
record,” Q38: “I am attending <SCHOOL> basketball games
lately because of the team’s national small college ranking,”
and Q39: “The team’s record does not really affect my attendance
level”) were not significant in explaining attendance in the original
model.

That identification with players (Team Familiarity) resulted in being
the most important factor is not surprising for a smaller college. For
current students, the chances of knowing a player are likely to be greater
at smaller colleges.

Based on this sample, encouraging connections to players (Factor 5: Team
Familiarity) and the college (College Affiliation), in that order, will
have the greatest impact on encouraging heavy use. The results suggest
the following guidelines, roughly in order of importance, for encouraging
heavy users in small college basketball. These suggestions should be viewed
as complementary to the findings of Wells et al. (2000). Although their
study involved small-college football and our study basketball, we suspect
the same would be true for other sports.

• Make team members accessible to fellow students and community
members. Do not have special dormitories, etc. which would separate student
athletes from fellow students. Also, encourage other participants in the
sporting event (e.g., cheerleaders, members of the pep band, etc.) to
interact with students and the community.
• Encourage identification of the community and students with the
college.
• Help potential fans understand basketball better in an attempt
to convert people to true fans. Sessions with coaches and players in which
past games are analyzed and current strategy is discussed might be helpful.
These sessions would help with the previous two bullets as well.
• Ensure awareness of the times and dates of games. Merely printing
a schedule is not enough. Market segments must be identified in terms
of how best to aggressively inform them of the times and dates.
• Schedule college events to avoid conflicts with the sports schedule.
• Use promotions and other activities to improve the excitement
and entertainment value of the sporting event, taking care to make sure
that these activities are complementary to the event and do not detract
from it.

Exhibit I

Likert Scale Variables and Highest Factor Loadings
(1 = Strongly Disagree, 5 = Strongly Agree)

Factor Labels:
Factor 1 = College Affiliation, Factor 2 = Entertainment, Factor 3 = Affiliation
with the Sport
Factor 4 = Time Constraints, Factor 5 = Team Familiarity, Factor 6 = Lack
of Awareness

Exhibit I (continued)

Factor Labels:
Factor 1 = College Affiliation, Factor 2 = Entertainment, Factor 3 = Affiliation
with the Sport
Factor 4 = Time Constraints, Factor 5 = Team Familiarity, Factor 6 = Lack
of Awareness

References

1. Baade, Robert A. and Laura J. Tiehen (1990). “An Analysis of Major League Baseball Attendance, 1969-1987.” Journal of Sport and Social Issues. v.14(1), pp.14-32.
2. Graham, Peter J. (1994). “Characteristics of Spectators Attending Professional Tennis Tournaments in Two Regions of the U.S.” Sports Marketing Quarterly, v.3(3), pp.38-44.
3. Graham, Stedman, Joe Jeff Goldblatt, and Lisa Delpy (I 995). In The Ultimate Guide to Sport Event Management & Marketing. Richard D. Irwin Inc. Helitzer, Melvin (1996). “In The Dream Job: Sports Publicity, Promotion and Marketing.”
4. University Sports Press. The E.W. Scripps School of Journalism, Ohio University. Athens, Ohio.
5. Hunt, Shelby D., Van R. Wood, & Lawrence B. Cainca (1998). “Corporate Ethical Values and Organizational Commitment in Marketing.” Journal of Marketing 53(July), pp.79-90.
6. Krohn, Franklin B. and Clarke, Mark (1998). “Psychological and Sociological Influences on Attendance At Small College Sporting Events.” In College Student Journal. v.32 (2), June. pp.277-287.
7. Lehnus, Darryl L. and Glenn A. Miller (1996). “The Status of Athletic Marketing in Division IA Universities.” in Sport Marketing Quarterly v.5 (3), pp. 31-48.
8. Mael, F. & B.E. Ashforth (1992). “Alumni and Their Alma Mater: A Partial Test of the Reformulated Model of Organizational Identification.” Journal of  Organizational Behavior Behavior, 13, pp.103-123.
9. Mawson, L. Marlene, and Edward d. Coan (1994). “Marketing Techniques Used by NBA Franchises to Promote Home Game Attendance.” Sport Marketing
10. Quarterly v.3(1), pp.37-45.
11. Murrell, Audrey J. and Beth Dietz (1992). “Fan Support of Sport Teams: The Effect of a Common Group Identity.” Journal of Sport & Exercise Psychology v.14, pp. 28-39.
12. Noll, Roger (1974). “Attendance and Price Setting.” in Government and the Sports Business. The Brookings Institute Washington, D.C. pp.115-157.
13. Parsons, Patricia H. (1993). “Framework for Analysis of Conflicting Loyalties” Public Relations Review. 19(l), pp.45-57.
14. Pitts, Brenda G. and David K. Stotlar (1996). In Fundamentals of Sport Marketing, Fitness Information Technology Inc., Morgantown, WV.
15. Pol, Louis G. and Sukgoo Pak (1994). “The Use of Two-Stage Survey Design for Gathering Data From People Who Attend Sporting Events.” Sport Marketing Quarterly v.3, pp. 9-12.
16. Sutton, William A., Mark A. McDonald & George R. Milne (1997). “Creating and Fostering Fan Identification in Professional Sports.” Sport Marketing Quarterly_ v. 6(1), pp.15-22.
17. Tajfel, H. (1981). in Human Groups and Social Categories: Studies in Social Psychology. Cambridge, England: Cambridge University Press.
18. Wakefield, Kirk (1995). “The Persuasive effects of Social Influence on Sporting Event Attendance.” Journal of Sports & Social Issues, v. 19 (4) November. pp.335-352.
19. Wann, Daniel L. and Michael P. Schrader (1997). “Team Identification and the Enjoyment of Watching A Sporting Event” in Perceptual and Motor Skills, pp. 84-954.
20. Wann, D.L. & N.R.A. Branscombe (1993).”Sports Fans: Measuring Degree of Identification With Their Team.” International Journal of Sport Psychology, v. 24, pp. 1-17.
21. Weeth, Charles (1994). “Fan Loyalty: Rose Bowl Boosts University of Wisconsin’s Revenue.” In Amusement Business. (Mar) v. 106 (10), pp. 7-13.
22. Wilkenson,D.(1993). In Sponsorship Marketing:A Practical Reference Guide for Corporation’s in the 1990 ‘s. The Wilkenson Group; Sunnydale, CA.
23. Zhang, James J., Dennis W. Smith, and Dale G. Pease (1996). “Spectator Knowledge of Hockey as a Significant Predictor of Game Attendance.” Sport  Marketing Quarterly v 5 (3), pp.41-48.