Table Manners: A Retrospective on the Online Poker Craze


By Marc A. Londo and Maria Cristina Santana

WJMCR 27 (December 2010)

Introduction | Social Cognitive Theory | Method | Results | Discussion


In late 2003, there was a surge in the popularity of tournament poker television shows on American television. The poker craze became a marketing phenomenon, and its influences were the cause of speculation in student newspapers across the country. This study was conducted near the apex of its popularity in early 2006. It was designed to measure the impact of the newly popular television genre on students. A sample of 444 college students across the country completed a survey designed to assess gratifications sought through media along with measures of attitudes, gambling behavior, and social systems. Using a Social Cognitive framework, exposure to these shows was measured and evaluated. The results indicated that student gambling was strongly correlated to viewership of poker shows, particularly among the youngest students utilizing the online gambling option. Gambling behavior of peers wasn’t shown to be a strong influence for student gambling. Excitement was shown to be a significant factor.

Keywords:Social Cognitive Theory; Activation Theory; College students/Risk taking; Youth/Gambling; Student life


Student gambling is a wide-spread social behavior on college campuses. However, for all the public conjecture surrounding it, very little is known about the causes of student gambling because there have been few studies addressing this issue.1 This lack of understanding is crucial because there is a variety of negative effects that have been recognized to accompany this behavior. These effects have been shown to include increased rates of suicide and attempted suicide,2 disruption of work and educational endeavors, criminal arrests and other legal issues,3 financial distress and familial disruption.4 There have also been studies drawing positive correlations between student gambling and alcohol/drug usage, anxiety, depression, eating disorders, and smoking.5
In 2003, a new variety of poker-themed television programming began to draw a significant portion of the college-aged demographic. These new shows blended aspects of the extremely popular reality television genre with coverage of poker tournaments. Their utilization of behind the scenes footage made the shows fresh because viewers gained an insiders view of the game while it was in progress. Many speculated that students were increasingly gambling on poker because they found the shows seductive. In an article appearing in USA Today, Columnist John Saraceno warned of the glitzy appeal of the shows.6 He compared the characters in the broadcasts to “cult-like pop figures,” while suggesting their glamorized images may be putting students at risk of becoming “the next inveterate gambler.” There was particular concern over the increased access that online gambling sites were giving underage gamblers.

According to Bandura’s Social Cognitive Theory,7 identification with media characters can lead to modeling through a reciprocal interaction between one’s behavior, environment, and personal influences. This study sought to ascertain the implications of this unique media phenomenon on college students through the application of Bandura’s model.

The Poker Phenomenon

In the summer of 2003, the debuts of ESPN’s “World Series of Poker” (WSOP) and the Travel Channel’s “World Poker Tour” (WPT) drew a record number of viewers. The staggering success of these shows turned heads since ESPN’s relationship with the WSOP went back 10 years, and they had never established such a following. Much of the success of the shows in 2003 was credited to Steve Lipscomb (Producer of the WPT) whose idea to imbed a miniature camera (called the “pocket cam”) into each player’s table position provided viewers with an insider’s perspective of the game. This feature effectively brought the audience into the game by revealing the hole cards of each player as they wagered increasing amounts on each hand. Another reason for the success of the ESPN show, which averaged 1,248,000 viewers in 2003 compared to 408,000 for the same time slot in 2002,8 was the feel good story of Chris Moneymaker. Moneymaker, an accountant from Tennessee, won his place in the tournament through an online poker website and outlasted 839 players en route to a $2.5 million grand prize. His amazing story inspired a new generation of amateur poker players that were honing their skills online. Many of those players would go on to become celebrities on the shows. The success of the amateurs on these shows perpetuated the idea that anyone could become rich by playing the game. In the process, the online poker industry quickly grew into a billion dollar enterprise.

Student Gambling

Between 4% and 8% of college students can be classified as pathological gamblers, with the rates for males being significantly higher than for females.9 The likelihood that college students will experience both pathological and problem gambling disorders is nearly double that of the general adult population.10 In 2002, Neighbors found that close to 15% of college students were at risk for gambling problems.11 This number was considerably larger than that for the general population, which runs between three and five percent.

A study by Platz and Millar12 discovered that college students were motivated to gamble largely for social reasons, which included exploration, being with friends, and being with similar people. These motivations were not all that different than their motives for engaging in other recreational activities such as watching television. The only motive that problem gamblers rated in their top five motives, for both recreational and gambling activity, was excitement.


Identification occurs when a viewer shares a media characters perspective and vicariously participates in their experiences while viewing. The ability to participate vicariously in the experiences of media characters, at times to the point of identity loss, is a recognized cognitive function that has implications on viewer perceptions. The camera establishes the point of view and is determinative of the target of audience identification. For those following the exploits of their favorite poker stars, the “pocket cam” provided an inside look at the inner workings of the game from the viewpoint of their favorite players.

This phenomenon has been observed in previous studies. Huesmann et al.13 found that identification with aggressive characters on television increased the learning of aggressive behaviors by children. In a study of celebrity endorsers, Basil14 found that identification with celebrities who were promoting health messages increased the adoption of those messages.

There may be numerous reasons behind a viewer’s identification with a particular player. Those reasons may reinforce their motives for playing the game. Consequently, it’s conceivable that a student who identifies with a certain poker personality may seek to emulate their gambling tendencies. In order to gain a more encompassing view of the affect of viewer identification, the following hypothesis and research question were posed for this study.

H1: Higher identification with the characters that are featured in the televised poker shows will be associated with increased levels of gambling behavior.

RQ1: Which gratifications sought from gambling are positively associated with higher degrees of identification with televised poker?

Identification with media prominently factors into the development of self-identity and social attitudes.15 Toy’s cognitive structure approach to the communication process, as seen in Figure 1, demonstrated how attitudes are determinative of behavioral intentions (B.I.), which are causally related to behavior, as moderated by situational factors.16

Figure 1 Toy’s Cognitive Structure Approach

Previous research has linked media usage (both instrumental and ritualized use) with media selection and attitudes.17 In studies regarding televised violence, researchers discovered a positive relationship between exposure and aggressive attitudes.18 The effect of attitudes on behavior has been studied extensively in advertising research. According to Toy’s cognitive structure approach, any behavioral change that manifests as a result of identification with poker shows should be accompanied by an attitudinal change. Thus, the following hypotheses were posed.

H2: Higher exposure to gambling programming on television will be associated with more positive attitudes regarding gambling.

H3: Higher identification with the characters that are featured in the televised pokershows will be associated with more positive attitudes regarding gambling.

Social Cognitive Theory

According to Social Cognitive Theory (SCT), identification with media others may produce modeled behavior. This change in behavior is produced from vicarious insight derived through glimpses obtained from media models. These glimpses can be powerful predictors of future behavior. SCT defines behavior as an observable act driven by expected outcomes. These expectations may be formed as a result of direct experience or mediated by vicarious reinforcement through others.

The triadic reciprocal model of SCT, as shown in Figure 2, provides a structure in which action (motor responses, verbal responses, & social interaction), personal factors (cognitive, physical, & affective), and environmental factors (social interaction), combine to influence behavior.19 Their influence is neither simultaneous nor equal in strength. The variance that each individual exhibits within this model is governed by five basic human capabilities: symbolizing capability, forethought, self-reflective capability, self-regulatory capability and vicarious capability.20

Figure 2 Bandura’s “Triadic Reciprocal Action”

Through the symbolizing capability, cognitive representations of anticipated future events act as incentives and motivators to re-enact those scenarios in the future.21 Their influences, both vicarious and real, are cognitively filtered. Consequently, they provide college students with the capability to learn, create and test gambling scenarios without actual participation. Media representations influence the ways in which the consequences are perceived by emphasizing positive rewards while limiting sanctions. Through forethought, individuals have the ability to devise future plans of gambling participation by weighing the consequences against the projected rewards.

According to Bandura,22 the self-reflective capability is a dimension of self-influence that allows individuals to reflect upon themselves and the adequacy of their thoughts and actions. Thus, it allows individuals to make comparisons between the value of a thought and how it measures up to reality. These comparisons can either be based on one’s beliefs and evaluations, common knowledge, perspective versus perspectives of others, and/or vicarious verifications, through any number of sources, including media.

The connection between identification and identity is especially important during adolescence.23 During adolescence, identification shifts from parents to peer groups and a more stable identity is formed. Peer groups play a vital role in identity formation through their influence on the choices of the individual. As that individual ages and moves on to college, these choices are further removed from the regulative capacity of the familial structure. Thus, it stands to reason that peer influence combined with increased access to gambling opportunities, could have a marked effect on gambling attitudes and behavior.

While certain aspects of the college community are known to predispose that population to gambling behavior, it is the reciprocal interaction of social and cognitive forces that direct the process of identity shaping as the student matures into adulthood. Peers have a strong influence over one another through their behavioral evaluations. It is through feedback that students further develop their sense of self. Positive feedback results in a greater degree of satisfaction. That parallel relationship is emblematic of the self-regulatory capability.24

Studies by Devlin and Peppard25 show that college students perceive their friends as having the highest rate of problem gambling. When considering that peer approval is so instrumental in the development of the self-regulatory capability, it stands to reason that the perceptions of peers may have a significant impact on the gambling habits of the individual. Consequently, if the perception is that “everyone is doing it,” a student may feel isolated from the group if they aren’t participating.

The media can implant ideas either directly or through social networks. They can also influence the adoption of trends by giving the impression that everyone else is doing it. Because students are continually making appraisals of their social aptitude, it’s important to consider the perceptions that surround them. To address the impact of peer gambling on the behavior of college students, the following hypotheses were formulated:

H4: Students that have friends who gamble with a greater frequency will have morepositive attitudes regarding gambling.

H5: Students that have friends who gamble with a greater frequency will be morelikely to gamble themselves.

When learning vicariously through mass media, viewers position themselves as learners. They pay close attention to the learned behavior and assess the outcomes that follow. Since most people interact with a small portion of the world, their perceptions about social reality are often shaped this way. According to Ball-Rokeach and DeFleur,26 as more people define reality through the mediated symbolic environment, the greater the social impact of the media.

A large amount of information regarding behavioral patterns is gained through modeling in the symbolic environment of mass media. SCT was developed to explain these findings, which Bandura couldn’t account for previously with Social Learning Theory. He was especially interested in identification with television personalities.27 In the past, Social Learning Theory (SLT) was effectively utilized in gambling studies involving adolescents who were shown to model the gambling behavior of their family members.28 Griffiths and Wood29 used Bandura’s SLT as an explanation for how the national lottery appeals to adolescents in the U.K. through their use of celebrities in the national media

Between 2003-2004 (the year the shows debuted), a survey by the Annenberg Public Policy Center at the University of Pennsylvania found an 84 percent increase in weekly card playing among young men between the ages of 14 and 22.30 Considering the developmental characteristics that accompany late adolescence (i.e., increased levels of stress, impulsivity, and depression), along with the influence of the college environment on the shaping of attitudes, there are justifiable concerns tied to the appeal of this product among this population. The following hypothesis was proposed to measure the impact of this programming on the gambling behavior of college students.

H6: Higher exposure to gambling programming on television will be associated withincreased levels of gambling behavior.

The Activation Effect

What made this phenomenon unique was the concurrent emergence of a highly successful gambling-themed television genre and expanding online gaming technologies. Previous media studies into the influence a single media can exert over the use of another media have shown a positive relationship.31 The “Activation Effect” occurs when there is an increase in the use of a particular form of media that results in a moderate supplemental increase in the use of another media. The ‘Cinderella story’ of internet qualifier Chris Moneymaker in 2003 triggered a dramatic rise in the number of online poker players in future broadcasts. Consequently, the success that these online poker players had in the shows was credited with sparking the online poker phenomenon.

Since these shows attracted a sizeable portion of the 18-25 year-old male demographic, the following hypothesis and research question were posed to assess the influence of these shows on the online gambling behavior of college students.

H7: Televised poker viewership will be positively associated with the level of online
gambling behavior.

RQ2: Which gratifications sought from televised poker is positively associated with
higher degrees of online gambling?

According to Griffiths and Wood,32 the increased access and availability to gambling provided by online sites has the potential to encourage excessive gambling behavior. In other countries, research evidence has shown that greater access to gambling leads to an increase not only in the number of regular gamblers but also in the number of problem gamblers.33 These effects are compounded by the interactive nature of the internet, which tends to make people feel less inhibited.34

College students are a high-risk population for experiencing both pathological and problem gambling disorders, with rates nearly double the general population of adults.35 Considering the intrinsic issues of increased access and lower inhibitions resulting from the solitary nature of online gambling, the following hypothesis was posed.
H8: Those students who display increasingly problematic gambling behavior will make greater use of the online gambling option.



Data were obtained from students at selected universities across the United States. Approximately 444 self-report questionnaires were obtained from a regionally diverse range of public universities that allow public access to student email address listings. The largest portion of respondents fell into the 22 years and older category (39.4%), 18.8% were 21 years old, 12.4% were 20 years old, 15.2% were 19 years old, and 14.3% were 18 years old. A total of 49.7% (n=219) of the respondents were male and 50.3% (n=222) of the respondents were female. Subjects were made aware that their participation would be voluntary and that their privacy would be protected.

The universities that were chosen represented a variety of population sizes and geographical characteristics. Of the universities represented in this sample, three were from the Pacific region (the University of California Berkeley, the University of Washington, and Washington State University), two were from the Southwest (Arizona State University and Colorado State University), two were from the Midwest (the University of Minnesota and the University of Missouri), two were from the Southeast (Clemson and West Virginia University), and one was from the Northeast (Rutgers, The State University of New Jersey).


The sample was generated from the online databases at the aforementioned universities. The email addresses of 22,800 college students were chosen at random and invitations were sent to their listed addresses inviting them to take part in the study. However, 2,082 invitations bounced and never reached their recipients. Thus, 20,718 solicitations were emailed to potential participants between Dec. 11, 2005 and Jan.16, 2006. Of those who received the invitation to participate, 669 viewed the survey website and 444 completed the instrument for a 2% response rate.

Each email address was tracked through QuestionPro’s ( automatic respondent tracking system. However, unless the student viewed the website, there was no way of tracking how many students actually viewed the survey invitation. Once the last group of respondents received invitations and filled out the instrument, the raw data were downloaded into an SPSS file for analysis.


Attitude. Strong et al.’s36 Gambling Attitudes and Beliefs Scale (GABS) was utilized to measure attitudes. The GABS was designed to predict gambling involvement among college students. Involvement of students is assessed through a set of 10 items that rank-order subjects according to their positive attitudes and beliefs about gambling.

Gratifications Sought. To gain a complete perspective of gratifications sought through televised poker, I employed measures of identification along with scales relative to media choice and choices in gambling behavior.

Identification. Two unidimensional measures of identification were used to rank the distance between the viewers and the television personalities.37 The specific questions were:

How much would you like to be like the players you see on TV?
Are there things that you see the players do that you would like to do?

The first question was rated using a 5-point scale, ranging from “not at all” to “exactly like them”. The following question was answered by using a simple “yes” or “no”.

Media Choice: The Television Viewing Motives Scale38 was used to assess the following motives for watching television: relaxation, companionship, habit, pass time, entertainment, social interaction, information, arousal, and escape. This 27-item measure was set on a 5-point Likert scale (1 =not at all; 5 =exactly) and was averaged to create an index of the motives for television viewing.

Choices in Gambling Behavior: Neighbors et al.39 polled college students on their motives for gambling and found that more than 70% of respondents endorsed money, enjoyment, social, excitement, and boredom. A 15-item measure, designed to measure expected gratifications from gambling behavior, was constructed using the reasons that were given as an example for each motive in the study as the basis for a 5-point Likert scale (1 =not at all; 5 =exactly). These items included; Money (e.g., “make money,” “win money,” and “get rich”); Enjoyment (e.g., “to have a good time,” “it’s enjoyable,” and “it’s fun”); Social (e.g., “social interaction,” “to be with friends,” and “to socialize”); Excitement (e.g., “for the rush,” “excitement,” and “it’s exciting”); and Boredom (e.g., “something to do,” “pass time,” and “bored”). They were then averaged to create an index of the motives for college student gambling.

Media Exposure: Each participant was asked to indicate the number of times that they watched poker shows on television within the last month. Their responses were scored using a 5-point scale, ranging from “never” to “5 or more times a week”.

Online Activity: Online activity was assessed by asking participants to estimate the number of hours they spend online playing poker per week.

Social System: To assess the effect of social relationships on gambling behavior, respondents were asked whether friends and family played poker and how frequently the friends and family played. An index of social influence was created by averaging the individual items.

Gambling Behavior: The gambling behavior of each respondent (online and social) was measured by asking two questions pertaining to their poker playing. These questions were, “how many hours do you spend online playing poker per week?”, and “how frequently do you play cards for money each month?” The answers to these questions were then added and average for a cumulative index of gambling behavior.

Problem Gambling Behavior:. The South Oaks Gambling Screen (SOGS) was used to measure problem gambling behavior. This 16-item device is widely used to screen individuals for pathological gambling or problem gambling behavior in the general population and clinical settings. Participants answered “yes” or “no” to initial questions regarding gambling behavior and, in later items, chose from a list of responses relating to frequency of gambling and amount of money spent on gambling. One point is assessed for each “yes” response. Responses were then summed. Scores of three or above were classified as problem gamblers, while those who score 5 or more were classified as probable pathological gamblers.40


Hypothesis 1: Higher identification with the characters that are featured in the televised poker shows will be associated with increased levels of gambling behavior. A Pearson correlation coefficient was used to determine the relationship between identification with the personalities that are seen in poker shows and gambling behavior. A positive correlation was found (r = .563, p<.001), denoting a statistically significant relationship between the two variables. Increased identification is positively correlated with higher levels of gambling behavior (social as well as online). Hypothesis 1 is supported.

Hypothesis 2: Higher exposure to gambling programming on television will be associated with more positive attitudes regarding gambling. A Pearson correlation coefficient was calculated for the relationship between exposure to poker shows and the Gambling Attitudes and Beliefs Scale. A positive correlation was found (r = .435, p<.001). The effect of exposure to gambling programming on attitudes toward gambling is statistically significant. Hypothesis 2 is supported.

Hypothesis 3: Higher identification with the characters that are featured in the televised poker shows will be associated with more positive attitudes regarding gambling. A Pearson correlation coefficient was utilized to calculate the relationship between identification with the personalities in poker shows and gambling attitudes. A positive correlation was found (r = .447, p<.001), denoting a statistically significant relationship between the two variables. Increased identification is positively correlated with more positive attitudes towards gambling. Hypothesis 3 is supported.

Hypothesis 4: Students that have friends who gamble with a greater frequency will have more positive attitudes regarding gambling. A Pearson correlation coefficient was used to calculate the relationship between perceptions of greater peer gambling behavior and attitudes towards gambling. A positive correlation was found (r = .251, p<.001), denoting a statistically significant relationship between the two variables. Attitudes toward gambling are positively correlated with perceptions of greater peer gambling. Hypothesis 4 is supported.

Hypothesis 5: Students that have friends who gamble with a greater frequency will be more likely to gamble themselves. A Pearson correlation coefficient was utilized to calculate the relationship between perceptions of friend gambling behavior and individual gambling behavior. A positive correlation was found (r = .337, p<.001), denoting a statistically significant relationship between the two variables. Friend gambling is positively correlated with an increase in individual gambling behavior. Hypothesis 5 is supported.

Hypothesis 6: Higher exposure to gambling programming on television will be associated with increased levels of gambling behavior. To calculate the relationship between exposure to gambling behavior and gambling behavior, a Pearson correlation coefficient was used. Table1 shows a positive correlation (r = .563, p<.001). The influence that exposure to gambling programming has on gambling behavior is statistically significant. Hypothesis 6 is supported.

Hypothesis 7: Televised poker viewership will be positively associated with the level of online gambling behavior. A Pearson correlation coefficient was used to calculate the relationship between poker viewership and online gambling behavior. A positive correlation was found (r = .461, p<.001). Poker viewership is shown to have a statistically significant influence on the level of online gambling behavior. Hypothesis 7 is supported.

Hypothesis 8: Those students who display increasingly problematic gambling behavior will make greater use of the online gambling option. A Pearson correlation coefficient was utilized to calculate the relationship between problem gambling behavior and online gambling behavior. A positive correlation was found (r = .442, p<.001), denoting a statistically significant relationship between the two variables. Problem gambling is positively correlated with increased online gambling activity. Hypothesis 8 is supported.

Research Question 1 asked which gratifications sought from gambling are positively associated with higher degrees of identification with televised poker. The strongest positive correlation was between identification and excitement (r = .528, p<001). Enjoyment (r = .512, p<.001) shared the next biggest correlation with identification. The correlation with boredom (r = .418, p<.001), money (r = .413, p<.001), and social (r = .400, P<.001), showed a weaker association with identification.

Research Question 2 asked which gratifications sought from televised poker were positively associated with higher degrees of online gambling. The strongest correlations that online gambling had with TV gratifications was with arousal (r = .390, p<001), information (r = .378, p<.001), entertainment (r = .370, p<.001), relaxation (r = .357, p<.001), and habit (r = .353, p<.001). Social interaction (r = .269, p<.001), escape (r = .251, p<.001), and companionship (r(409) = .243, p<.001) showed weaker, but significant positive correlations.

The Target Demographic

Analysis of the data suggests varying levels of receptivity among college students to poker programming. Nevertheless, when looking at this data, it’s important to remember that correlations define associations, not causation. Behavior may affect viewing or viewing may affect behavior. The import of these data resides in its ability to provide a greater perspective on gambling media and how it manifested within this demographic. Considering the diversity of the college population, it’s important to understand the relationships that may promote risky behavior. Thus, it’s useful to view this phenomenon according to a variety of demographical factors.

Gender. Correlation analysis indicates that for college males exposure to poker programming shares significant positive correlations with hours spent playing poker online (r = .493, P<.001), frequency of social card playing (r = .542, P<.001), identification (r = .588, P<.001), gambling attitudes (r = .478, P<.001), and gambling behavior (r = .596, P<.001). Those are significant differences from the entire population, which shows the following correlations with exposure: hours spent playing poker online (r = .461, P<.001), frequency of social card playing (r = .516, P<.001), identification (r = .620, P<.001), gambling attitudes (r = .435, P<.001), and gambling behavior (r = .563, P<.001).

For females, the same correlations were significantly weaker regarding exposure: hours spent playing poker online (r = .210, p=.002), frequency of social card playing (r = .293, p<.001), identification (r = .558, p<.001), gambling attitudes (r = .261, p<.001), and gambling behavior (r = .339, p<.001).

Age: Table 1 shows the strength of the relationships identification shares with exposure, online behavior, and overall gambling behavior, according to age. Since the correlation between identification to poker shows and online poker playing was so much stronger among 18 year olds, a question was raised as to whether Internet poker was being played more among younger students.

Table 1: Correlation Between Identification and Exposure, Online Gambling and Gambling Behavior According to Age







Exposure to Poker Programming

n = 58
r = .640

n = 62
r = .551

n = 51
r = .582

n = 77
r = .631

n = 157
r = .620

Online Poker Playing

n = 58
r = .668

n = 62
r = .460

n = 51
r = .514

n = 77
r = .265

n = 156
r = .337

Gambling Behavior

n = 58
r = .721

n = 62
r = .604

n = 51
r = .486

n = 77
r = .432

n = 156
r = .508







To gauge the differences in online poker playing according to age, independent-samples t tests were used to evaluate the differences in online poker playing between 18 year old college students and older college students. The results in Table 2 confirm that younger college students are spending more time online playing poker.

Table 2: T-tests Show Differences in Online Poker playing Between 18 Yr Olds and OlderStudents.








N = 61

N = 66

N = 54

N = 80

N = 169

M = 1.49

m = 1.318

m = 1.259

m = 1.10

m = 1.18

t = 2.992

t = 1.21

t = 1.54

t = 3.43

t = 2.436

p = .03

p = .012

p < .001

p < .001





Grade Point Average: The detrimental effect of increased and excessive gambling among college students (especially the younger students) is far too complex to be gauged by a simply survey. The consequences can be immediate as well as long term and they may affect both life circumstances and emotional health. Nevertheless, academic performance is a relevant measure for this population. A correlation analysis showed negative relationships between GPA and poker viewership (r = -.120), identification (r= -.079), gambling attitudes (r=-.136), and gambling behavior (r=-.126).

Social System: Previous studies have reported on increased gambling behavior within college student segments. Groups that have typically been reported as susceptible to increased gambling behavior have included athletes, Greeks, and students that live in campus housing. Since these university sponsored student groups make up a substantive portion of the social systems surrounding college life, any of these groups could be included in the environmental factors that serve in Bandura’s triadic model of reciprocal interaction.

Student housing is one such environment in which interaction between students has a profound effect on the social norms. It is quite common for there to be regular poker games in certain dorms/apartments. It is just as common for neighbors to gather around the television to catch their favorite shows. For those that watch poker, correlation analysis shows a strong positive relationship between poker watching and identification (r = .638), indicating a strong positive linear relationship between the two variables.

Considering the increased opportunities to participate in a poker game, it may not be surprising if those who identified with the shows would seek out a poker game. Among those that live in student housing, identification showed a strong correlation with both social card playing (r = .608), as well as online poker playing (r = .612).

The South Oaks Gambling Screen. According to our sample, 11% of males and 2% of females were at the problem gambler level, while 6.3% of all college students were classified as problem gamblers.

The Excitement Factor

Through analyzing the correlations across gratifications scales (poker viewership gratifications along with student gambling gratifications) the correlation between watching poker for arousal, and gambling for excitement, was significantly stronger than the relationships between the other gratifications (r = .629, p<.001). That this cross-correlation between gambling and television gratifications was so strong, raised questions as to whether these measures were drawing on a common factor(s) and, if so, whether the combining of these two measures into an overall index of excitement would provide a greater insight into how excitement influences the other motivations across measures.

In a study comparing the motives for gambling with other recreational activities, Platz and Millar41 found that problem gamblers cited excitement as one of their top five motives for participating in gambling and other recreational activities. This supports assertions made by Lesieur42 and others that some pathological gamblers are "action seekers" who don’t gamble for the money, but rather for the excitement associated with being in the action. Roy and Linnoila43 attributed the connection between excitement and gambling to a biological need, due to low levels of norepinephrine. This chemical of the brain is secreted under stress, arousal, and excitement. Therefore, pathological gamblers may engage in such activities as gambling to increase their levels of norepinephrine.

According to the American Psychiatric Association, late adolescents are highly vulnerable to depression because their biochemistry sometimes causes "deficiencies in two chemicals in the brain, serotonin and norepinephrine, which are thought to be responsible for certain symptoms of depression, including anxiety, irritability, and fatigue." When put in the context of Roy and Linnoila’s44 study on excitement and considering the various studies focused on students, depression,45 and pathological gambling,46 one must question whether there is certain excitement factor underlying the surface of this issue.

Combining the two excitement gratifications provided an overall excitement measure that showed a strong relationship between identification with the characters in poker shows (r(363) = .677, p<.001) and the Gambling Attitudes and Beliefs Scale (r(354) = .674, p<.001).

When comparing the correlations across scales, the excitement factor (with a Cronbach’s alpha of .90) showed strong correlations with student gambling gratifications: Money (r = .627 , p<.001), Enjoyment (r = .807 , p<.001), Social Interaction (r = .591 , p<.001), and Boredom (r = .595 , p<.001). Among poker viewing gratifications, this factor registered strongly significant correlations with relaxation (r = .682 , p<.001), habit (r = .724 , p<.001), pass time (r = .529 , p<.001), entertainment (r = .800 , p<.001), social interaction (r = .573 , p<.001), information (r = .717 , p<.001), and escape (r = .605 , p<.001). There was also a moderate correlation with companionship (r = .344 , p<.001).

Not surprisingly, this measure of overall excitement was significantly correlated with frequency of poker viewing (r = .642, p<.001). Poker viewing was strongest among males viewing poker shows (r = .651, p<.001), while among females this factor showed a moderately weaker correlation (r = .542, p<.001). Also, consistent with earlier age-related findings, the excitement factor was negatively correlated to age (r = -.143, p=.006) and classification (r = -.156, p=.003).

For the question “What’s the most you ever gambled in one day?” this factor showed a moderately strong significant correlation (r = .503, p<.001). There was also a moderately strong significant correlation with social card playing (r(367) = .554, p<.001).


In 2005, it was reported in the Chicago Tribune that the average daily number of players in the most frequented online poker rooms had increased from close to 2,400 in 2003 to about 60,000.47 This study was conducted not long after those figures were released and 10 months prior to the establishment of the Federal Unlawful Internet Gambling Enforcement Act in October of 2006. Consequently, it provides an insightful glimpse into the convergence of two emergent media phenomena and their combined influence on late adolescents.

What emerged from this research was a profile of the students who were being most affected by the poker explosion. A significant proportion of them were young, male, living in student housing, watching poker, playing poker online and identifying strongly with the stars they saw on television. Since this pattern of poker consumption was a relatively new phenomenon when many of the upper classmen were already attending classes, it’s not surprising that the youngest students showed the greatest effects in this sample. The troubling issue is how those effects manifest over time.

Gambling behavior is affected by a complex range of variables, some of which aren’t addressed by this study. According to Bandura, our actions are a product of a reciprocal interaction between our environment and our personal factors (cognitive, physical, and affective). The influence those factors exert on each other are neither simultaneous nor equal in strength. One of the more surprising observations in this study was that student gambling wasn’t more affected by peer and family gambling habits. Considering that gambling begins for many as a social exercise, it seemed likely that cohorts would have a greater impact on the gambling habits of their friends. However, it appears as though there are certain personal factors that play a greater role in deciding who is most affected in their gambling behavior. These personal factors are reflected through ones beliefs and evaluations which are determinative of their gratifications sought.

This study was fashioned around the basic principles of Social Cognitive Theory to test the impact of the increasingly popular poker television shows on student gambling. It was hypothesized that through identification with the personalities that compete in the poker shows, college students would experience a change in attitude that would manifest in increased gambling behavior. This attitude change would be reinforced and made stronger through interaction with their social networks. While the influence of social systems was shown to have a slight-significant impact on the model, identification was shown to have a significant impact on attitude, which in turn had a moderate-to-strong significant impact on gambling behavior. These effects were most strongly felt among the younger students. Most interesting was the finding that online poker playing was affecting the younger students to a much greater degree than their older counterparts.

The research questions were equally revealing. Through studying the television and gambling gratifications (and how they impact identification, problem gambling behavior, and online gambling), the excitement/arousal gratification showed strong significance across the two scales. When aligned into one excitement measure, they revealed a strong significance with many of the other gratifications along both measures. This common factor, which is inherent in the excitement measure, showed equally strong significance in the GABS scale as well as the identification index. Considering the extensive literature on excitement as a primary element of gambling addiction, the fact that it resonated so strongly in this study raises questions as to whether there is a biological component involved among college students, particularly among younger students with an under-developed identity.

By emphasizing social systems as a precipitator of student gambling, less attention was given the individual measures of beliefs and evaluations. The college environment is comprised of a diverse set of young minds. With so much diversity, more questions regarding personality may have provided a more detailed profile of those being most affected by the poker craze. Nevertheless, with 88 questions in the instrument, it just wasn’t possible to incorporate other measures.

While there were some concerns with using student email as a way of distributing the survey, the rewards justified the costs. It allowed for a diverse sample group that wasn’t limited by geographic proximity. Data collection and data entry were both more efficient and didn’t require random checks to see that the forms were being entered in properly. Using a website made filling out an 88 item survey much less stressful on the respondents. Considering the number of variables in the instrument, it is a positive that two-thirds of the students that logged on to the site took the time to complete the survey.

Marc A. Londo is a Ph.D. candidate at Temple.  Mara Cristina Santana is an associate professor of communication and director of women’s studies at University of Central Florida.  This was one of the top five papers in the Entertainment Studies Division at the 2006 convention of the Association for Education in Journalism and Mass Communication.

Leave a comment

Your email address will not be published. Required fields are marked *