Gambling Disposition In College Students
Sarah P. Belter
Christy L. Palbicki
Dennis N. Lorenz
University of WisconsinGreen Bay
Gambling is the fastest growing addiction in the United States (Horn, 1997), and the college campus is not immune to this trend. The college experience often includes an increased risk for developing gambling habits (Winters, 2001), and consequently, more college students suffer from gambling addictions than others in this age group (Lesieur, et. al., 1991). In one survey, 33 percent of males and 15 percent of females in college gamble at least once per week (Saum, 1999), and in another survey, 84-91 percent of college students polled had gambled at least once in the prior year (Winters, 2001). Even students who do not gamble are faced with greater pressure than ever to place their bets (Shaffer, Hall, Vander Bilt, & George, 2003). Academic institutions across the country are challenged with the growing issue of problem gambling by students (Shaffer, Forman, Scanlan, & Smith, 2000).
Part of the attraction is related to the excitement surrounding the game, winning money, social interactions, (Neighbors, Lostutter,, Cronce, & Larimer, 2002), and sports events. Students like to bet on their home teams, NCAA events, and professional sports (Saum, 1999). Moreover, student athletes are twice as likely to be problem gamblers as non-athletes (Saum, 1999), plus competitive students, male and female, score higher on intrinsic and extrinsic motivations for gambling revealed through the Gambling Motivation Scale (Burger, Dahlgren, & McDonald, 2006).
Another reason why college students gamble is the quick opportunity to pay bills. Students from a small Canadian community college were asked to list the motivation for their gambling. Their top response (80%) was the possibility of winning money. This is consistent with the idea that many college students are turning to gambling as a quick fix to financial problems (Adebayo, 1998).
A problem that gamblers of all ages face is the lack of good judgment during the excitement of the bet. They may be attentive to minute details and scores of statistics, but when it comes to betting versus rational thinking, gamblers often go with their intuition. Consequently, the results can be devastating because many of their ideas are irrational (Delfabbro & Winefield, 2000). One such irrational premise that contributes to the demise of many gamblers is the gambler’s fallacy, also called negative recency, or maturity of the chances (Ladoucer & Dube, 1997; Lepley, 1963; Ross & Levy, 1958). Despite the fact that the statistical probability remains constant for certain games of chance like coin flipping and roulette, individuals affected by the fallacy feel that the odds will soon change favoring one side of the coin over the other. For example, if the outcome of 5 coin flips was HHHHH, the bettor gets a strong gut feeling that the next result will be tails because the average outcome in this event will be 50/50 before long. The fallacy has two parts: the first is an historic attribution—the coin has a memory of how it landed in the past; and the second is that nature will soon correct any chance event that goes too far in one direction by presenting the opposite outcome a sufficient number of times to offset recent trends.
Another problem many gamblers face is referred to as the gambler’s dilemma or chasing (Shaffer, & Hall, 1996). This occurs when a gambler is losing and believes by increasing the wager that s/he will quickly recover loses. This strategy may be successful on occasion, but frequently leads to larger bets and greater losses.
In an attempt to better understand the potential for gambling in college students, the phenomenon of maturity of the chances was investigated along with the effect of winning or chasing in the current experiment. Subjects were also questioned regarding their gambling habits.
Methods
Participants
Undergraduate students (364) from a small state university in the Midwest participated in our study. The students were randomly assigned to one of four groups, 91 per group. The experiment, in progress over a 12-month period, was conducted in a classroom with an overhead projector and screen positioned at the front for the presenters. Participation was voluntary, and most students were given a small amount of class credit. Our university IRB approved the project.
After the students were seated in the classroom, one investigator read the opening statement welcoming everyone and requesting the participants to keep the context of the experiment confidential. All were informed that they would not be penalized if they did not want to answer the question, or felt the need to leave the experiment. Subsequently, no one left the experiment or failed to respond during any of the trials.
The investigator then instructed the participants that the answers would be either yes (tails) or no (heads), or blank (no response). Each group responded to a unique set of situations presented on the overhead screen while one investigator read aloud the three scenarios. All were asked to respond to each scenario before hearing/seeing the next. There were four groups, each given a different set of coin flip situations. The independent variables included the number of prior head outcomes and their order, and whether the student was winning or losing money. No coins were tossed in this experiment. The dependent variables included the number of tails, heads, or blanks generated by each given scenario, and responses to a brief survey following the coin-flip questions.
Group 1: Scenario 1—“You are at the casino. You are currently $35.00 in the hole. You are betting on a coin-flip where the prior outcome has been H,H,H,H,H. The wager is at $50.00. Would you bet your money on a tails outcome (or heads, or nothing were the other options)? Scenario 2—“You are at the casino… the prior outcome has been H,H,H. Scenario 3—“You are at the casino… the prior outcome has been H.
Group 2: Scenario 1—“You are at the casino. You are currently $35.00 ahead. You are betting on a coin flip where the prior outcome has been H,H,H,H,H. The wager is at $50.00. Would you bet your money on a tails outcome (or heads, or nothing were the other options)? Scenario 2—“You are at the casino... the prior outcome has been H,H,H. Scenario 3—“You are at the casino… the prior outcome has been H.
Group 3: The position of being “ahead” was the same as Group 2 except the order of presentation was reversed (H first then HHH followed by HHHHH).
Group 4: The position of being “in the hole” was the same as Group 1 except the order of presentation was reversed (H first then HHH followed by HHHHH).
Finally, participants were asked to complete a brief survey about gambling habits following the experiment. Everyone was asked two questions: “Would you consider yourself a gambler?” and “How much money would you typically spend at the casino?” Their responses were stapled to their data sheet to ensure a matched set. Chi-Square statistics were used to analyze the non-parametric results.
Results
The present experiment revealed a number of significant betting behaviors and dispositions toward gambling held by college students who, for the most part, claimed not to be gamblers.
Group 1:
The results for group 1 revealed that students who were losing (down $35) bet on tails as a function of the history of coin-flipping events that were presented in descending order (5H, 3H, 1H; see Fig. 1). When told the coin previously landed on heads five times in a row, students bet the next flip would produce a tails outcome more often than when told the previous history brought about just one heads X2(1, N=92)=3.522, p=0.061. The differences between the other coin-flip histories revealed a predictable trend, but were not statistically significant (5H versus 3H X2(1, N=103)=0.476, p=0.490, and 3H versus 1H, X2 (1, N=85)=1.424, p=0.233).
Group 2:
In contrast to Group 1, Group 2 was winning (up $35) when they placed their bets, nonetheless, the results for Group 2 revealed a similar trend (see Fig. 2). When told the coin previously landed on heads five times in a row, students bet the next flip would produce a tails outcome more often than when told the previous history brought about just one heads X2(1, N=86)=3.767, p=0.052). The differences between the other coin-flip histories followed a predictable trend, but were not significant (ie. 5H versus 3H, X2(1, N=98)=0.367, p=0.544), and 3H versus 1H, X2 (1, N=80)=1.800, p=0.180).
Group 3:
Groups 3 and 4 were presented with the ascending scenario of coin-flips (1H, 3H, 5H). The results for Group 3 revealed that students who were wining (up $35) also bet on tails as a function of the history of coin-flipping events (see Fig. 3). As the history of heads increased from 1H to 3H, more students believed that tails would occur (1H versus 3H, X2(1, N=93)=4.742, p=0.029). Progressing from 3H to 5H in the scenario also increased the betting on tails, but not significantly (3H versus 5H, X2(1, N=119)=0.210, p=0.647). The largest preference for tails was evident when comparing bets after the 5H versus the 1H X2(1, N=98)=6.898, p=0.009). Note that the preference for betting on heads (60%) in the beginning of this experiment has been demonstrated in a previous experiment involving college students (Lepley, 1963).
Group 4:
Group 4 was losing (down $35) but bet on tails in a pattern almost identical to Group 3 that was up by $35 (see Fig. 4). As the history of heads increased from 1H to 3H, more students believed that tails would occur (1H versus 3H, X2(1, N=94)=2.085, p=0.149). Progressing from 3H to 5H in the scenario also increased the betting on tails, but not significantly (3H versus 5H, X2(1, N=116)=0.552, p=0.458). The largest preference for tails was evident when comparing bets after the 5H versus the 1H X2(1, N=102)=4.745, p=0.029).
There was no order effect overall when comparing betting patterns. No significant differences appeared during the ascending progression of heads when compared with the descending progression. Statistical analysis of groups 3 and 4 where the history of coin-flips went from 1H to 3H to 5H, versus subjects in groups 1 and 2 where coin-flips went from 5H to 3H to 1H revealed a trend to bet on tails while ascending, but the results were not statistically significant (gps 3+4 vs. 1+2 at 1H, X2(1, N=147)=0.170, p=0.680); (gps 3+4 vs. 1+2 at 3H,), X2(1, N=205)=1.410, p=0.235); and (gps 3+4 vs. 1+2 at 5H, X2(1, N=231)=1.251, p=0.263). This indicates that the forward order of presentation (1H, 3H, 5H) did not elicit a significantly greater number of tail bets than the reverse presentation (5H, 3H, 1H), regardless of whether the subjects were wining or chasing.
Subjects who were down $35 (groups 1 & 4) did not bet more often on tails than those who were up by $35 (groups 2 & 3), X2(1, N=583 for 3 measures)=0.139, p=0.709).
We also asked each subject two personal questions about gambling. The majority of subjects in all four groups claimed they were not gamblers, and they spent very little at the casino. This raises an interesting contradiction. All of our subjects, supposedly non-gamblers, bet on every scenario without a single refusal.
In Summary, college students in our experiment revealed a strong penchant for negative recency—betting on tails in the face of multiple heads results. The negative recency effect was in full force when students were presented with the 5H scenario. Overall, the ascending scenarios elicited slightly more bets on tails than the descending scenarios, but there was no statistical difference for this variable. There was a slightly preference to bet on tails while chasing in this experiment, but again, no statistical difference for this variable.
Discussion
The main result was that college students were eager to bet (non-existent money) on pre-determined coin-toss situations, and were influenced by the history of coin-flips. Independent of the ascending order and level of winning, they consistently bet significantly more often on tails after 5H than 1H. Apparently more students felt tails were “due” (negative recency) rather than betting with the consecutive heads “on a role” (positive recency). No one questioned whether the coin or flipping technique was legitimate.
Our college students showed a significant preference toward the negative recency effect in runs as few as 3H during the ascending sequence of 1H, 3H, 5H. The positive recency effect, betting heads “on a roll” was not significant at any time, nor could it be distinguished from random guessing or other reasons why a minority of subjects chose heads. The positive recency effect may take more trials or gambling experience before it is preferred over the negative recency effect (Lindman & Edwards, 1961).
Do gamblers suspend the rules of logic and enter into the realm of “maturity of the chances” if they believe a pattern is emerging in a game of chance? In a previous experiment, winning as few as four times in a row was enough for many college students to assume they had control over the coin flip situation (Langer & Roth, 1975). Also winning early in a series of coin flips gave many college students the illusion of control (Langer & Roth, 1975; Lepley, 1965). A more recent experiment suggests that frequent messages reminding subjects of the notion of randomness in the gambling situation significantly reduced their strength of erroneous beliefs and reduced the number of trials played (Benhsain, K., Taillefer, A. & Ladouceur, R., 2004). This reinforces the idea that gamblers, swept away with the excitement of the game, may benefit from reminders about the actual odds of the game.
One hundred percent of our subjects bet on every scenario. The result may have been related to the fact that they were not actually wagering any money, or possibly because there is a greater interest in gambling than the average student is willing to admit. It is difficult to create actual gambling conditions in a controlled situation in which subjects are willing to bet their own money. However, the purpose of the present experiment was not an attempt to reproduce a casino setting, but to assess gambling disposition in college students, a population whose pathological gambling level is estimated to be between 3.4 percent and 5.9 percent (Shaffer, Hall, & Vander Bilt, 1999). The coin flip test described above is not likely to be a good predictor of problem gambling in college students because the negative recency effect was so prevalent in our sample. Nonetheless, information regarding the dominant and rapid occurrence of negative recency over positive recency may be useful developmental landmark to help school administrators and councilors in their assessment of gambling cognition, and consequently to develop Gambling Awareness Programs tailored to meet the needs of their own institutions.
The past does not predict the future in fair chance events. Yet, intermittent reinforcement for correct guesses creates positive feelings of success and control (Langer & Roth, 1975). The number of incorrect guesses tends to fade in the light of a few correct responses, especially if winning occurs in the beginning (Langer & Roth, 1975). Grade school students, middle school and younger, tend not to be influenced by negative recency, but tenth graders and adults show a strong preference for the phenomenon, especially after a run of four identical events like 4H (Ross & Levy, 1958). A major factor has emerged since this seminal report by Ross and Levy. Young students today are part of the first generation in this country to be surrounded by legalized gambling (Shaffer, et.al. 2003). The onset of gamblers’ fallacy in grade school students deserves to be revisited.
Regardless of the of the developmental issues we are still grappling with the question: Can high school and college students learn to recognize the difference between a gut feeling and the true odds for simple events of chance? There is some evidence that younger students can benefit from a gambling awareness program (Ferland, Ladouceur, & Vitaro, 2002), but there is no information regarding the success of a similar program for college students (Shaffer, et. al., 2000). Ultimately, students at all levels and university officials will be faced with the challenge of recognizing the transition from gaming for recreation, to self-destructive involvement in gambling addiction (Shaffer, Donato, LaBrie, Kidman, & LaPlante, 2005).
The human interest/ability to categorize and predict, enhanced by brain areas of reward to reinforce correct choices, makes gamblers of us all. One does not need a casino player’s card or a regular lottery habit, anyone who has automobile or homeowners’ insurance, life insurance, or a mutual fund is playing the gambling game. Government sponsored lottery programs, gaming contracts with Native Americans, and Internet casinos give tacit approval to gambling at many levels (Collins, 2003; Barker & Britz, 2000). Given the fact that gambling in it sundry forms is a part of life, how long will it take before universities add Gambling 101 to their general education program, and what curriculum will best serve their needs?
References
Adebayo, B. (1998). Luck of the dice: Gambling attitudes of a sample of community college students. College Student
Journal, June, Vol. 32, Issue 2.
Barker, T., & Britz, M. (2000). Jokers Wild. Westport, CT: Praeger Publishers.
Benhsain, K., Taillefer, A. & Ladouceur, R. (2003). Awareness of independence of events and erroneous perceptions
while gambling. Addictive Behaviors, 29(2), 399-404.
Berger, T. D., Dahlgren, D., & McDonald, C. D. (2007). College students and gambling: An examination of gender
differences in motivation for participation. College Student Journal 40(3), 704-714.
Collins. (2003). Gambling and The Public Interest. Westport, CT: Praeger Publishers.
Delfabbro, P. H. & Winefield, A. H. (2000). Predictors of irrational thinking in regular slot machine gamblers.
Journal of Psychology, 134(2), 117-129.
Ferland, F., Ladouceur, R., & Vitaro, F. (2002). Prevention of problem gambling: Modifying misconceptions and
increasing knowledge. Journal of Gambling Studies, 18(1), 19-29.
Horn, B. P. (1997). Is there a cure for America’s gambling addiction? USA Today magazine. May, Vol. 125,
Issue 2624, 34.
Ladouceur, R. & Dube, D. (1997). Erroneous perceptions in generating random sequences: Identification and strength
of a basic misconception in gambling behavior. Swiss Journal of Psychology - Schweizerische Zeitschrift fuer
Psychologie – Revue Suisse de psychologie. 56(4), 256-259.
Langer, E. J. & Roth, J. (1975). Heads I Win, Tails It’s Chance: The Illusion of Control as a Function of the Sequence
of Outcomes in a Purely Chance Task. Journal of Personality and Social Psychology, 32(6), 951-955.
Lepley, W. M. (1963). “The Maturity of the Chances”: A Gambler’s Fallacy. The Journal of Psychology, 56, 69-72.
Lesieur, H. R., Cross, J., Frank, M., Welch, M., White, C. M., Rubenstien, et al. (1991). Gambling and Pathological
Gambling among University Students. Addictive Behaviors 16, 517-527.
Lindman, H. & Edwards, W. (1961). Supplementary report: Unlearning the Gambler’s Fallacy. Journal of Experimental
Psychology, 62(6), 630.
Neighbors, C., Lostutter, T. W., Cronce, J. M., and Larimer, M. E. (2002). Exploring college student gambling motivation.
Journal of Gambling Studies, 18(4), 361-370.
Ross, B. M. & Levy, N. (1958). Patterned predictions of chance events by children and adults. Psychological Reports,
1, 87-124.
Saum, W. S. (1999). Sports gambling in college: Cracking down on illegal betting. USA Today Magazine. July,
Vol. 128, Issue 2650, 62.
Shaffer, H. J., Donato, A. N., LaBrie, R. A., Kidman, R. C., and LaPlante, D. A. (2005). The epidemiology of college
alcohol and gambling policies. Harm Reduction Journal, 2(1), 1-20.
Shaffer, H. J., Forman, D. P., Scanlan, K. M., Smith, F. (2000). Awareness of gambling-related problems, policies and
educational programs among high school and college administrators. Journal of Gambling Studies, 16(1), 93-101.
Shaffer, H. J., and Hall, M. N. (1996). Estimating the Prevalence of Adolescent Gambling Disorders: A Quantitative
Synthesis and Guide toward Standard Gambling Nomenclature. Journal of Gambling Studies, 12(2): 193-214.
Shaffer, H. J., Hall, M. N., Vander Bilt, J. (1999). Estimating the prevalence of disordered gambling behavior in the
United States and Canada: A research synthesis. American Journal of Public Health: 89, 1369-76.
Shaffer, H. J., Hall, M. N., Vander Bilt, J., and George, E. M. (2003). Futures at Stake: Youth, Gambling, and Society
. Reno, NV: University of Nevada Press.
Winters, K. (2001). Gambling and college students. Minnesota Institute of Public Health. Gambling Problems Resource
Center. http://www.miph.org/gambling/gmb_collegestud.html
Legends
Figure 1. Number of bets on heads and tails during the descending progression of 5H to 1H while subjects were $35 in the hole.
Figure 2. Number of bets on heads and tails during the descending progression of 5H to 1H while subjects were $35 ahead.
Figure 3. Number of bets on heads and tails during the ascending progression of 1H to 5H while subjects were $35 ahead.
Figure 4. Number of bets on heads and tails during the ascending progression of 1H to 5H while subjects were $35 in the hole.
|
