Are College Athletes Psyching Themselves Out?

Study suggests perceived threat of confirming negative stereotypes may cause some -- but not all -- of academic achievement gap between athletes and non-athletes.
February 11, 2009

More than a year ago, a major study of 71 Division III member institutions of the National Collegiate Athletic Association documented a significant academic achievement gap separating male athletes and non-athletes at selective liberal arts colleges. Although admissions practices -- potentially biased in favor of enrolling athletes -- might be to blame for this gap, a new study suggests that the perceived threat among athletes of confirming the negative stereotype of the “dumb jock” might also help perpetuate the gap.

A new “working paper,” released this week by the National Bureau of Economic Research, explores a phenomenon known as “stereotype threat” among athletes at Swarthmore College. Thomas S. Dee, the paper’s author and Swarthmore economics professor, writes that this trend “refers to the perceived risk of confirming, through one’s behavior or outcomes, negative stereotypes that are held about one’s social identity.” Dee argues that the “stigma” attached to athletic participation at some selective institutions might trigger the “stereotype threat” response among athletes, accounting for some portion of their weaker academic performance.

The papers follows the spirit of earlier research by the College Sports Project -- an Andrew W. Mellon Foundation study of Division III athletics -- by focusing on the “social identity” of athletes at “selective colleges and universities,” where the author claims there is a growing “athletic-academic divide.” Through a controlled experiment, Dee’s research details the notable extent to which “stereotype threat” compromised the academic performance of athletes at his institution.

“It’s fair to say that there is some history of conflict between the academic and athletic culture here on campus,” Dee said of Swarthmore. “There was a lot of anxiety created, in particular, when the college eliminated football nine years ago. That catalyzed a lot of raw feelings on both sides of the issue. More recently, there have been conflicts over scheduling priorities. Given this, I think it’s fair to say there is a stereotype in the community about the role and ability of student-athletes.”

The Experiment

Last spring, all Swarthmore students were presented with the opportunity to participate in a lab experiment, simply labeled, “to examine the determinants of cognitive functioning” with no hint of its athletic focus. To bring in more athletes, follow-up requests to the initial invitation were made specifically to them. Ultimately, 84 students participated in an hour-long study. Of those students, 44 percent of them were athletes -- a sample greater than the college-wide proportion of 24 percent. Otherwise, the study population “closely” resembled the student population.

Students, athlete and non-athlete alike, were then grouped into a “treatment condition” and a control group. Both groups were given the same set of 39 questions, taken from the quantitative and verbal sections of the Graduate Record Examination. The former group was “primed” before taking this test by being given a brief questionnaire on basic information which included the question, “Are you (or have you been) a member of a National Collegiate Athletic Association sports team at the college?” These students where then asked with what frequency they had experienced any scheduling conflicts -- on a scale from 1 to 7 -- due to their involvement in athletics. The control group received the same questionnaire minus the questions about being an athlete and instead were asked about dormitory conditions.

After being given 30 minutes to answer the GRE questions, participants in both groups were given a short word-completion exercise to test what Dee calls the “cognitive activation" -- or self recognition -- of the implied stereotype presented to the "treated" group. Finally, all participants were given a closing questionnaire, and here the members of the control group were finally asked if they were athletes.

The Results

For all students, the average score on the test was about 74 percent. Those athletes given the “stereotype-threat treatment” -- meaning that their status as an athlete was identified prior to being given the test -- earned test scores significantly lower than those who did not. On average, this “treatment” reduced the scores of athletes by 11 percentage points, 14 percent below the average for non-athletes.

Dee believes this outcome confirms his hypothesis that some of the academic achievement gap between athletes and non-athletes -- at least at his institution -- can be explained by athletes psyching themselves into below-average performance. Although some may doubt the relevance of these findings outside of the intense scrutiny of the lab setting, Dee said he believes the nature of his experiment closely mimics the rigors of the real-world setting of a college campus.

John D. Emerson, principal investigator at the College Sports Project and mathematics professor at Middlebury College, said he agrees with Dee that the “stereotype threat” could be a plausible explanation for this performance gap. Still, he cautions readers not to assume it is the only cause. He argues that different standards of admission for athletes and non-athletes at these institutions more than likely account for the majority of this gap.

“I would be disappointed if people at colleges -- actually on the ground where the rubber meets the road -- interpret this hypothesis and say, ‘It’s something else. It’s not what’s going on in the admission office and letting in people with different or lower academic credentials. It’s this thing called stereotype threat,’ ” Emerson said. “And believe me; they’re not going to understand what that really means, by the way. I hope people do not think this removes the challenge from admissions. That would be unfortunate.”


Closing the Gap

While acknowledging the limited application of his findings, Dee did argue they could be used to field further research and possibly influence policy recommendations. He suggested that athletes could potentially be “buffered” from the effect of the “stereotype threat.” In his model, an academic adviser or coach would sit down with athletes at the beginning of the academic year and inform them of this phenomenon.

“People call it forewarning,” Dee said of this ability to combat the threat. “Simply warning people about a phenomenon can disarm the feeling that they don’t belong. As a practical matter, you may be able to overcome the stereotype threat by frequently making use of this method.”

He said he would like to see a group like the NCAA implement a wider study of this method to research its practical viability. Such a large case study could be implemented on 12 or so randomly assigned campuses, he said, adding that this would ultimately provide enough research to provide explicit guidance on how to replicate this “buffering” elsewhere if it is successful.

A concerted effort to reduce “stereotype threat” on the campuses of selective liberal arts institutions, Dee said, could help reduce the performance gap between athletes and non-athletes. Emerson went one step further, noting that only when institutions truly blindly admit athletes will this gap be closed.


Back to Top