The subjects of the study are students, but there was no attempt to randomize the sampling. Instead, the authors chose the schools and the athletic programs within the schools. The students self-selected for participation, which also may have affected the participation rates. Because the schools and athletic programs were involved, the subjects may have been wary of the promise of confidentiality, causing the results to have some bias. Singer, von Thurn and Miller (1995) note that when the data is sensitive, confidentiality assurances can improve the quality of responses. However, the confidentiality promise needs to come from the researcher in order to be completely trustworthy. Thus, the methodology introduced some bias into the study with the design of its confidentiality mechanism, and the researchers introduced bias when they failed to randomize school and program selection. The independent variable is the prevalence of drug usage. The dependent variables were the NCAA division, the team, the gender and the race/ethnicity of the respondents. The responses were in the form of categorical variables (yes or no). The responses were subject to Chi-squared analysis to "determine if sport division or ethnic category could differentiate drug-use patterns among collegiate student-athletes" (p.52). This test does provide information regarding the statistical difference between the groups (p.52) but the authors choose not to interpret the data further. The authors discuss the statistical, but not practical significance...

This approach is appropriate, since the role of the study is to inform policy, not set policy.
The issue with the paper lies not with the methods of statistical analysis but rather with the methods of collection. The study design's major flaw is the lack of randomization and the lack of true confidentiality. Given that the sensitivity of the information that the students were asked to provide, better assurances of confidentiality, gained from direct interaction with the researchers, would have been more ethically sound and also would have delivered more accurate response. Further, the lack of randomization could lead to researcher bias interfering with the reliability of the data. This bias risk is not addressed in the paper; this is an oversight on the part of the researchers. Unfortunately, the reliability of the results suffers from these deficiencies, and the study should be read with an understanding of the methodological flaws in mind.

Works Cited:

Green, G., Uryasz, F., Petr, T. & Bray, C. (2001). NCAA study of substance use and abuse habits of college student athletes. Clinical Journal of Sport Medicine. Vol. 11 (2001) 51-56.

Hart, C. (1999). Doing a literature review: Releasing the social science research imagination. SAGE Publications, Inc.

Singer, E., von Thurn, D. & Miller, E. (1995). Confidentiality assurances and response a quantitative review of the experimental literature. Public Opinion Quarterly. Vol. 59 (1)…