Unbiased? Race, gender, and sport effects in university medical staff’s perceptions of injured student-athletes

ABSTRACT

Injured student-athletes rely on university medical staff personnel for care. Do these practitioners exhibit race and/or gender biases in their perceptions of injured student-athletes? While such biases have been widely documented in other medical practitioner populations, they have not been studied in the domain of college athletics. We use a survey experiment conducted on National Collegiate Athletic Association Division 1 medical staff to explore perceptions of an injured student-athlete (e.g. the likelihood of the student-athlete complying with treatment). We find little evidence of bias. We discuss why this population of medical practitioners may differ from others, and we offer suggestions for future work on medical treatment of student-athletes.

KEYWORDS:

• Sport injuries
• NCAA medical staff
• race bias
• gender bias
• medical treatment
• ACL injuries
• survey experiment
• student-athletes

Acknowledgement

We thank David Figlio, Alexander Green, Adam Howat, Kevan Ketterling, and Jennifer Richeson for helpful advice. We thank Alexandria Fredendall, Noah Kanter, and Allison Rubenstein for research assistance (in helping to gather sample information). There is no conflict of interest involved.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 See the appendix for a detailed rationale of design specifics as well as some pre-test results regarding the vignette.

2 Due to random assignment, we are in a strong position to make causal inferences about the impact of race, gender, and sport since, on average, the groups assigned to different vignettes should differ only in terms of variations in what their vignette describes.

3 The mean score, on a 4-point scale (with higher scores indicating the need for more restrictions), for women soccer players is 3.23 (std. dev. = .69; N = 155). The mean score for all others is 3.07 (.70; 494) (t₆₄₇ = 2.58; p ≤ .01 for a two-tailed test).

4 The average combined percentages exceed 100% since we did not confine respondents to come to a 100% total.