The Extent and Nature of Gambling Among College Student Athletes


o Table of Contents
o Executive Summary
o Introduction
o Review of the Literature
o Methodology
o Results
o Discussion
o Conclusion
o Bibliography
o Biography
o Acknowledgements
o Contact Information
o Purchase Copy


This research study had two primary purposes. The first was to assess the types of gambling behavior in which college student athletes engage. The second was to determine the rates at which student athletes participate in these activities and, more specifically, to see if patterns of participation differed by gender. A self-administered, written questionnaire was used to gather responses to closed-ended questions. Approval to use human subjects for this investigation was obtained from the University of Michigan Institutional Review Board. The specific research methods are described in the following sections.

Setting for the Study Selection of participants, which were student athletes from football, and men's and women's basketball teams, was as follows:

  1. All NCAA Division I-A football playing schools were included in the study.
  2. After identifying 112 Division I-A institutions, the institution's football and basketball conference was reviewed with the goal of including all of the members of the basketball conference in the population. For example, Temple University competes in football in the Big East and in the Atlantic 10 in basketball. Therefore every school from the Atlantic-10 was included in the study.
  3. If an institution was included in the sample because one of the institution's conference members plays Division I-A football, their football program (if one exists) was included in the sample. For example, Villanova University was included in the study because their basketball program competes in the Big East. Their football athletes, even though they compete at the I-AA level, were also included in the sample of football athletes. However, Villanova's football conference was not sampled. Institutions from the Ohio Valley Conference and Big Sky Conference were added to round out the sample.
  4. The institutions from 17 athletic conferences were included in the sample. The complete list of conferences included: Atlantic Coast (ACC), Atlantic-10, Big 12, Big East, Big Ten, Big Sky, Big West, Colonial, Conference USA, Mid-American (MAC), Ohio Valley, Pac 10, Patriot League, Southeastern (SEC), Sun Belt, Trans American Athletic (TAAC), and Western Athletic (WAC).
  5. A total of 182 institutions were represented in the sample. All of the institutions competed in men's and women's basketball. Only 148 of the institutions competed in football.
  6. Participants were selected randomly from each athletic conference.
  7. The final sample of 3000 student athletes consisted of 1500 football players, 750 men's basketball players, and 750 women's basketball players.


Data Collection Procedures

Participants' names and addresses were obtained from team rosters and information available through various public sources (e.g., institutional directories, World Wide Web sites, and public phone records). The randomization process resulted in approximately 10% of each institution's football team (assuming approximate roster sizes of 105 participants) and 27% of each institution's men's and women's basketball team (assuming approximate roster sizes of 15 participants) being included in the sample.

Via regular mail, each participant received a cover letter that described the purpose of the study, explained the confidentiality protections used, and offered information on how to contact the researchers with any questions. In addition, the packet included a self-administered questionnaire, pencil, and a stamped reply envelope addressed to the investigators. No payment or compensation was provided to respondents since such payments would have violated the NCAA's rules against providing extra benefits to student athletes.

It is important to note that neither the survey, nor the return envelopes, were coded in any manner. The omission of any type of identifying marks was done in order to increase the confidence of the respondents that the research was truly anonymous. Because the surveys were not coded, there was no possibility of sending a second instrument to those individuals who may not have responded to the first query.


Survey Instrument

The questionnaire was developed using concepts and items taken from three previously published studies "The Extent and Sources of NCAA Rules Violations" (Cullen and Latessa, 1996), "The South Oaks Gambling Screen" (Lesieur and Blume, 1987), and "A Survey of American Gambling Attitudes and Behavior" (Kallick, Suits, Dielman, and Hybels, 1979). The current survey instrument incorporated two questions from Cullen and Latessa's research. The majority of our survey instrument was adapted for a student athlete population from The South Oaks Gambling Screen with most of the response items being retained. However, some items that screen for problem and pathological gambling were not included in the final instrument. Finally, Kallick, et al was used to refine the survey instrument as it pertained to a student athlete population and to create more specificity in identifying potential types of gambling behavior. The questionnaire was pilot-tested four times prior to the national distribution to assure an easily understood survey instrument.

The questionnaire was divided into four sections: (I) general gambling activities; (II) sports related gambling with friends; (III) gambling with bookmakers and other organized gambling activities; and (IV) demographic and other general information about the respondent. Each of these sections will be described in turn.

Section I began with a series of questions about specific gambling behaviors and asked respondents to indicate if they engaged in the behavior since attending college and the number of times in the last 12 months. Among the gambling behaviors queried in this section were playing cards for money, betting on animals or sports, and gambling in casinos. Additional questions probed the amount of money gambled since attending college; borrowing money; missing classes, practices, or competitions due to gambling; and reasons for gambling.

Sections II and III addressed issues specifically related to sports gambling. Section II was designed to measure the number of times and amount wagered on bets against friends on specific sports and the reasons for it. Section III addressed specific sports betting, typical bets, and reasons for betting with bookmakers. Respondents were also asked if they ever bet on a game in which they participated, received money for not playing well, or provided inside information about a game to aid bookmakers.

Section IV included standard questions about age, sex, race, family income, and year in school. Other questions sought information about athletic eligibility, living situation, grade point average, sport(s) played, and whether or not the respondent received an athletic scholarship. Respondents were also asked to assess their role on their team.


Data Preparation and Analysis

Careful review of all of the data to correct any data entry errors and remove any partially completed survey information was conducted prior to the data analysis. The primary purpose of this research was to provide descriptive information regarding student athlete gambling. Where appropriate, ANOVA was used to determine significant differences between male and female student athlete responses. The findings are presented below.


Table of Contents | Executive Summary | Introduction | Review of the Literature | Methodology | Results
Discussion | Conclusion | Bibliography | Biography | Acknowledgements | Contact Information