The findings in our report suggest greater involvement in sports-related gambling activities than earlier research.
The baseline data provided by Cullen and Latessa (1996), finding that 25% of student athletes in their sample gambled on college sports, appears to underreport the extent of sports-related gambling that is occurring nationwide because of their use of a more narrow definition of gambling. Most notable is our finding that over 45% of male athletes reported gambling on sports since attending college.
The increase may not necessarily signal an increase in gambling, but rather use of a survey instrument designed to more accurately measure student athlete gambling. The exact wording of Cullen and Latessa's question regarding sports gambling (which provides the benchmark for today's gambling discussions) was "While you have been in college (including other colleges you may have attended), have you gambled money on other college sporting events?" This question is limiting because it excludes wagering on professional sports. Also, it does not suggest to the respondent what gambling on college sports might include, such as sports pools, or wagering with a bookmaker or friend. Our research suggests that gambling on sports by student athletes is a prevalent activity.
There are additional findings related to the issue of sports gambling that should concern athletic administrators nationwide, including:
There are three limitations that are worth noting in this research:
A few factors may have contributed to under-reported gambling activities. Our highest proportion of returns was among female student athletes, which was also the group that had the lowest rates of gambling. This may indicate that the most likely individuals to return the survey were those student athletes that did not gamble. Concerns about being caught in an activity that could affect a student athlete's eligibility or about "social desirability" may have further resulted in an underreporting of gambling behavior. This possibility gains additional plausibility when one considers that 80% of male student athletes engaged in some type of gambling behavior but fewer than 25% of men's basketball athletes returned the survey.
It is also possible that the data over-reported student athlete involvement. Individuals who had no gambling involvement may have discarded the survey because they believed it was not relevant to them. This seems to be a less likely scenario in light of our high female response and their lesser gambling involvement.
While a high response rate is always desirable, the researchers were satisfied with the response rate for a number of reasons. First, no tracking method was used to increase the response rate. By comparison, Cullen and Latessa (1996) had a slightly higher response rate of 32.4%, but student athletes in their sample were sent at least two (and in some cases three) questionnaires as well as follow up letters. Considering the nature and sensitivity of the topic, a conscious decision was made to trade a lower response rate for a stronger guarantee of confidentiality for the participants. Second, NCAA regulations make offering incentives to student athletes to take the survey impermissible. Third, the nature of the topic may have caused some individuals to not return the survey due to perceived threats to their athletic eligibility.
Finally, some items from the South Oaks Gambling Screen that are concerned with problem and pathological gambling were eliminated from the survey instrument. While the SOGS is the recognized instrument for determining these types of gambling, our goal was not to establish a measure of these activities. Rather, our goal was to expand upon the findings of Cullen and Latessa and confirm or refute the gambling rates that they found in their earlier research. We believe the instrument we used accomplished this goal.
Directions for Future Research and Practice
It is the intention of the researchers to develop additional information about student athletes and gambling from the data that were collected. Potential correlates such as year in school, race, drug or alcohol use, team role, and differences between sports are just a few of the possibilities. The data are currently being reviewed so that the current study may represent a beginning, rather than an end, of current research about student athlete gambling. Analyses of this information and other correlates will be forthcoming.
Additional research is needed in a number of areas related to student athlete gambling. Our recommendations for future research include:
The NCAA has recognized the problem with student athletes and gambling and has done a commendable job in addressing the issue. However, additional steps need to be taken, which include: