A new study of 77 Division III institutions of the National Collegiate Athletic Association reveals a consistent and widening academic performance gap between athletes and non-athletes.

Monday, the College Sports Project – an initiative of the Andrew W. Mellon Foundation – released its second annual report as part of a five-year longitudinal study comparing the academic performance of athletes to that of non-athletes at participating Division III institutions. The research project has gained much attention because Division III, unlike Divisions I and II, does not track the academic performance of athletes separately from that of the rest of an institution’s student body.

Though NCAA officials argue that the data do not offer a fully representative cross section of the division, as the report favors more selective institutions, researchers from the Project believe it provides participating institutions a quantifiable and valuable way to track the academic outcomes of their athletes. The project does not take a prescriptive role or seek to police institutional behavior. Instead, its officials argue that tracking the outcomes of athletes will help participating colleges and universities meet their institutional goals.

After the first full year of college for the cohort of students who enrolled in 2006-7, male athletes, on average,had class ranks 9 percentile points lower than male non-athletes. (A gap between female athletes and non-athletes was present, but the report called it "modest.") Subdividing male athletes further, those who were recruited had class ranks 6 percentile points lower than those who were not recruited.

Looking at how the 2005-6 cohort of students fared by the end of their second full year of college, the report acknowledges that the gap in average class rank between athletes and non-athletes shrank by one percentile point. Though small, the report muses that this reduction could possibly indicate “that athletes gradually make positive adjustments to the demands of academic life.” Otherwise, the data show two phenomena continuing from last year’s report and mirroring this report’s new incoming cohort: female athletes uniformly did better than their male counterparts, and both male and female non-recruited athletes had grade point averages only slightly below those of non-athletes.

Percentile Class Rank GPA after Two Years: 2005-06 Entering Cohort Student Group

New to this year’s report is a breakdown of the data based on institutional selectivity. The GPA gap between athletes and non-athletes is greater at “highly selective institutions” -- those with an average combined SAT score of greater than 1250. At “less selective” institutions – those with an average combined SAT score of less than 1150 – the report found “no academic underperformance” by athletes. Research was also broken down by individual sport, although this and institution-specific information will not be publicly released by the Project’s coordinators.

Analyzing the Numbers

John D. Emerson, principal investigator of the College Sports Project and mathematics professor at Middlebury College, said there could be any number of reasons for the differences in college GPA between athletes and non-athletes. So far in the longitudinal study, he said, SAT score, high school grade point average, gender, race and ethnicity “do not fully account for the differences in college performance.” He also noted that the “time athletes spend on their sport” does not account for this either.

“Stereotype threat” -- or the theory that athletes underperform because they fear confirming the image of the “dumb jock” -- might account for some of the gap, Emerson said. The bulk of the cause, he said, is probably attributable to factors that the report’s regression model cannot measure.

“A real part of this is likely attributableto differences between groups of students at the point at which they are admitted to college, but differences that aren’t measured by high school test scores and grades,” said Emerson, careful to note that the data do not imply that some of these athletes were not prepared to attend or be admitted to these select institutions. “It may have to do with interest, and it may have to do with priorities or values. Another possibility could be differences during the college experience itself between the broadly construed culture of athletics and the culture of the rest of the student body.”

Similarly, he admitted there are no easy ways to determine why there is less of a performance gap between female athletes and non-athletes than male athletes and non-athletes. Still, he noted that the gap for women is growing and moving in the direction of that for men.

“Maybe some institutional priorities are such that people are willing to sustain greater differences between athletes and non-athletes in highly competitive sports for males than females,” Emerson said. “Maybe with sports like football and men’s hockey and baseball -- sports with a long history and tendency to draw bigger audiences and competition -- there is an incentive to give more emphasis to competition.”

NCAA Response

Though officials from the NCAA express their appreciation for the project and the spirit of its research, they caution that those who review the data in aggregate should not make broad assumptions about the division.

“As for the academic underperformance that’s documented, it’s important to realize that for this study only a small subset of Division III schools participated,” said Dan Dutcher, vice president of the division, which has about 450 members. “It’s not appropriate to assume those things exist across Division III. Still, I think the College Sports Project has served a valuable function for the division by raising the consciousness of the academic performance issue.”

The issue has gained such prominence in the division’s governance that the NCAA is considering implementing a limited reporting structure among Division III members -- comparing athletes to non-athletes in a method similar to the College Sports Project. The Division III Presidents Council initially mentioned the idea in a report last September, and Dutcher said it might come to fruition in the next two years. Still, he noted that some institutions do not see the need for this type of reporting, citing the division’s ban on any type of athletically related financial aid.

“The assumption has always been, for those institutions that don’t provide financial aid, that we should assume that you’re accepting student-athletes like other students and that they’re performing similarly,” Dutcher said. “Institutional autonomy is the other side of the coin. A lot of institutions have said, ‘What problem are you trying to solve? Because we don’t think that there’s a problem out there that justifies an administrative response.’ Some do not want to embark on a process if it creates an administrative burden.”

If and when an academic reporting structure comes to Division III, Dutcher said it would not have punitive measures similar to the Academic Progress Rate system that is used to hold Division I members accountable. He said he did not think it was likely that Division III would ever have initial or continuing eligibility requirements strictly managed and followed by the NCAA, like the other two divisions do. Instead, much in the spirit of the College Sports Project, any reporting structure would be used to inform institutions about the academic state of their athletes so they can take appropriate action.