Bias research art

Inspired by AdobeStock/anna42f
VIN News Service art by Tamara Rees

Claims of bias in veterinary medical school admissions have generally been the product of intuition, anecdote or extrapolation from research involving other professional schools. Now, a report from the Association of American Veterinary Medical Colleges puts meat on the bones of those conjectures with evidence of bias in the 2019 application cycle. 

Based on an analysis of survey responses from veterinary school applicants, the researchers reported scores of what they identified as statistically significant instances in which certain groups were disadvantaged directly or indirectly by the admissions process.

The disadvantaged groups were racial and ethnic groups underrepresented in veterinary medicine, applicants with limited financial means, women, first-generation college students, and those who grew up in rural communities and/or aspire to practice in rural communities.

"I have seen threads of these phenomena for many years, yet there had not been a big-picture view of how bias may be impacting accessibility within the applicant pool," Lisa Greenhill, AAVMC senior director for institutional research and diversity, said by email to the VIN News Service. "I am not surprised by the data, but I am saddened by just how many layers of experience and identity can make one's journey to the profession more challenging."

The AAVMC routinely conducts surveys "to explore the experiences of veterinary applicants" and has also done surveys focused on financial literacy and access to pre-veterinary advising. The new study marks the first time AAVMC used surveys of veterinary school applicants to explore the possibility of bias in the admissions process, Greenhill said.

"Admissions process" refers to the evaluation of candidates based on the application submission requirements, including coursework, test scores, experiential hours, essays, personal statements, recommendations and interviews, according to Greenhill, who co-authored the report. "This research examines how application requirements introduce bias in the preparation phase of applicants' development," she said. "The resulting bias then is introduced into the admissions process during application review."

The report comes during a period of heightened attention to the lack of diversity in veterinary medicine. Fueled by nationwide protests over systemic racism, the profession last year began taking a deeper and more widespread look at the causes and consequences of failing to represent the diverse society it serves. In a country where six in 10 people are white, nine out of 10 veterinarians are white.

"I am hopeful that the paper encourages the colleges to interrogate their approach to admissions," Greenhill said. "I am also hopeful that for the broader veterinary community, this paper helps to illuminate the challenges faced by applicants from marginalized backgrounds."

The Multicultural Veterinary Medical Association has called lack of diversity the most critical issue facing the profession. Dr. Marie Sato Quicksall, the group's president and a founding board member, said the bias report provides evidence that is difficult to dismiss. She praised the research for bringing deeper issues to light.

"Bias and discrimination is not always on the surface level. These deeper, less-seen issues are harder to address, and getting buy-in from people who are in the majority group to address hidden bias is also harder," she said. "It's easy to point to things like racial slurs and get support for combating the issue, but things like decreased access … can be a tougher sell."

Some veterinarians say the evidence would be more compelling if the study met more rigorous research standards. Dr. Mark Rishniw, research director at the Veterinary Information Network, an online community for the profession and parent of the VIN News Service, highlighted that the report is not peer-reviewed, nor did the authors make the data available for re-examination and additional analysis.

"The rule is to be as transparent as possible with the data," Rishniw said. "Seeing the tabulated or graphical results would help. Summary statements are insufficient."

Dr. Phillip Nelson, dean of Western University of Health Sciences College of Veterinary Medicine in California, said the report "gives us some reliable data to determine whether our suspicions are correct, or whether we're going in the right direction or not."

He was disappointed, however, that the report doesn't present data from the overall applicant pool. AAVMC operates the Veterinary Medical College Application Service (VMCAS), a platform used by 46 member schools in the U.S., Canada, U.K., Australia and the Caribbean. The association collects and reports annual trends using VMCAS data.

"If they could just make a statement that said that these demographics are similar to the overall applicant pool," Nelson said, "then I'd have a whole lot more security with addressing the results."

Asked about this criticism, Greenhill said the demographics are indeed similar. "[T]he data meets confidence intervals of 95-99%; we are confident that the data reflects the pool in its entirety."

'Disproportionate degrees of difficulty'

The bias report grew out of an earlier AAVMC study evaluating whether the quality of veterinary students overall was diminished as more students were accepted, due to larger class sizes and the addition of new schools. That study found that the number of academically qualified students substantially exceeded the number of admissions offers, and therefore concluded that the quality and depth of applicants as a whole remained strong. However, study author Dr. James Lloyd warned that his analysis of the demographic variables "strongly suggested" unintentional selection bias in the admissions process. He called for a specific examination of bias, which resulted in this latest effort. 

Lloyd, a veterinarian, economist and former dean of the University of Florida College of Veterinary Medicine, is a consultant who co-authored the bias study.

All veterinary school candidates using the VMCAS were surveyed twice. The first time was just after the close of the 2019 application deadline on Sept. 15, 2018. Of 8,152 surveys sent by email, 2,431 applicants responded, or 29.8%. A second survey went out after April 15, 2019, when those with acceptances were required to commit to a school or release their seat. That time, of 8,136 surveys sent, 1,958 responded, a rate of 24.1%.

Among broad findings of the survey: Black and Hispanic candidates were underrepresented in the candidate pool compared with their proportions in the U.S. population. Hispanic candidates were less likely than their Asian, white or Black counterparts to receive one or more offers of admission. Asians received at least one acceptance at a higher rate than any other group, including whites.

No significant differences or trends in the probability of receiving an offer of admission were found to be directly associated with a candidate's gender identity. However, women appeared to be directly or indirectly disadvantaged by the current admissions process, a striking finding since women account for a supermajority of veterinary students.

"Large numbers of women in the pool and in the enrolled classes should not be interpreted as the population being advantaged," Greenhill said. "The reality is that despite men now comprising roughly 12% of the applicant pool, they represent 18% of enrolled students (women currently comprise 82% of the DVM student population). On its face, this is indicative of a greater proportion of men being admitted based on their relative representation in the pool."

The report states that male applicants are more likely to have non-academic, pre-veterinary experiences that include an easier time getting experiential hours, more direct contact with veterinarians they shadow, and less difficulty gaining research experience.

"When coupled with general competitiveness across the board for all genders, such experiences position male candidates for more favorable admissions decisions," Greenhill said.

While race/ethnicity and gender identity were the primary dimensions of diversity examined for potential bias, the study examined secondary characteristics, as well, including:

• the number of times and the number of schools candidates applied to;
• their preparation in terms of experience hours and education;
• their socioeconomic status as indicated by whether they had received a Pell Grant (a federal subsidy for undergraduate students with extreme financial need) and their estimated current educational debt;
• whether they grew up in an urban, suburban or rural community; and
• where they hoped to practice after graduation.

Candidate confidence and concerns about admission, employment and finances also were assessed.

The authors posit that the primary and secondary factors alike affected the possibility of receiving an offer, and that their impacts sometimes compounded.

One example of the compounding effect, they said, is seen in candidates' education years before they apply to veterinary school: Advanced Placement classes in high school were associated with an increased likelihood of an admissions offer. Access to AP classes was less likely for first-generation students, Pell Grant recipients and students from rural communities. Higher proportions of candidates from underrepresented ethnic and racial groups are also first-generation students and/or Pell Grant recipients.

"[T]he barriers/deterrents are not absolute — many disadvantaged candidates are ultimately successful in gaining an offer for admission — but the playing field is certainly not level for all candidates; candidates from disadvantaged groups must overcome disproportionate degrees of difficulty to achieve their goals," Greenhill and Lloyd write. "What's more, to the extent that this situation might be recognized by potential applicants, it could actually be limiting the scope of possible candidates who might be willing to take on the challenging admissions process as they consider the prospects of a career in veterinary medicine."

The authors hope the study provides fuel for a "broad-based reexamination" of the admissions process, such as requirements for experience hours, among other potential barriers.

'Very powerful data'

Veterinary college deans contacted by the VIN News Service lauded the study as an important contribution to the work of making admissions more equitable.

"The facts speak for themselves," Dr. Paul Lunn, dean at North Carolina State University College of Veterinary Medicine, said. "Everything here supports our understanding. I think it's the strongest evidence you could hope to generate with the available data."

Dr. Rustin Moore, dean of the Ohio State University College of Veterinary Medicine, said, “This study helped reinforce why we are taking a deep dive into our admissions process to try and mitigate any bias facing our minoritized students and other applicants."

Even before the report came out, Lunn said NC State's veterinary school, like others, had taken steps to adjust its admissions criteria with the aim of eliminating bias. "The school has changed the GPA threshold to harmonize with contemporary standards and reduced the required number of veterinary and animal experience hours to make admissions standards more attainable," he said. "For the 2021 admissions cycle, the school suspended the requirement for the GRE graduate school entry exam because of the pandemic, and is considering whether there is any value in reinstating it for the future."

Lunn added that dialing back some criteria and heeding the AAVMC report are just early steps.

“Changing the prerequisites is a great way to increase access and help a more diverse group of candidates get access into the qualified pool. That's a positive,” Lunn said. “Whether it goes far enough is quite debatable.”

He said he wished the AAVMC study also had been able to look at admissions committees’ decision-making processes, an area in which NC State is focusing its attention.

As with many schools, the committee uses a "holistic review" process, Lunn said. "We have a number of agendas in doing that, but amongst the most important are meeting a goal to recruit a class that reflects our society, and have a process that is just and equitable." Committee members are provided training in inherent bias, using AAVMC guidance. It has also enlarged the admissions committee to include more members from outside the college, which Lunn said has broadened the diversity of opinions.

Looking ahead, Lunn said, the school is exploring doing more to make committee members blind to certain statistics, such as GPA and experience hours, "so that [candidates] are being considered on the merits of their application and references and what they represent in society."

Greenhill and Lloyd do reference admissions committees in the report. They cite a 2017 study of implicit bias in medical school admissions that found only a fraction of committee members who were made aware of their biases said they applied that awareness in making application decisions. "This suggests that while training and awareness can support greater equity in admissions processes, actual process changes may be needed to reduce the influence of bias in applicant evaluations," they write.

A closer look at experience requirements

Moore, dean at The Ohio State, said that while the AAVMC report found strong evidence of bias, it did not appear to explore the full scope of factors that can affect diversity, "such as gender identity and expression, ability, age and being a parent, that are often primary or intersecting areas of bias that can impact admittance and success. It does serve," he added, "as an important foundational study for further exploration of additional questions regarding factors impacting admission.”

Greenhill said the surveys did include questions about sexual identity and that the number of applicants identifying as LGBTQ is small. "Despite the small sample size, statistical testing was still completed," Greenhill said. "An analysis of this data showed no differentials based on this demographic and thus, there was nothing to report. As this population continues to disclose and/or grow amongst our application and student populations, we hope the case numbers will allow us to conduct more rigorous testing."

One compelling question the study raises, Moore said, is: What experiences are needed to be successful in veterinary school? 

The researchers gave considerable attention to requirements for veterinary, animal and research experience — requirements that many programs have dialed back in recent years over concern that they are inherently discriminatory. 

Nevertheless, the perception persists among applicants that accumulating as many experience hours as possible is important. The AAVMC surveys found that those with a greater number of animal-experience hours reported significantly greater confidence in their prospects of getting into veterinary school.

However, the survey found no significant relationship between an applicant's number of veterinary- and animal-experience hours and the probability of admission. Where significant differences were found was in the effort required to accumulate experience hours. Women, Pell Grant recipients and those belonging to racial or ethnic groups that are underrepresented in veterinary medicine reported greater difficulty obtaining veterinary experience. Women and underrepresented candidates also were less likely to spend their veterinary-experience hours with a veterinarian. Underrepresented candidates rated the experience lower in overall quality.

Those in underrepresented groups reported significantly fewer animal-experience hours and, along with first-generation candidates, reported greater difficulty in obtaining these hours.

Moore said he believes "there is no magic number of hours or type of experience" that qualifies a candidate to be admitted or that predicts who will be successful in school or the profession.

"Veterinary students and veterinarians come from all backgrounds and lived experiences, and we must consider any barriers or hurdles that could unintentionally prevent a very qualified person from gaining admittance to veterinary school and being successful in their dream and career aspiration …" he said.

Nelson, the Western University dean, found the information in the report about experience hours intriguing in a different way. What caught his attention is that respondents overwhelmingly described veterinary experience as "critical to their understanding of the veterinary medical profession."

He said the experience requirement has been long debated. One argument against it is: Other professional schools don't have in-the-trenches experience as a prerequisite; why should veterinary schools?

"What was enlightening to me was that the students themselves valued this experience," he said. "[They] realized that after finding out what veterinary medicine was all about, they still want to be veterinarians. So from my perspective, I have changed my position a little."

He also described as helpful the finding of no correlation between the number of veterinary and animal experience hours and getting an offer of admission. "It says that, apparently, we believe that there's a minimum number of hours you should have, and then after that, it's good," he said.

Still, the difficulty of getting hours is something "we have to pick apart," he added. "And I think that this analysis is going to help us do that."

Selecting for aristocrats?

While Nelson said the study raised more questions than it answered, he added that that's what good studies do. "It would be malpractice if we did not look at [the report], and then look at our program," he said.

One thing he'll be looking into, he added, is a worrying pattern that veterinary schools are "selectively working against those who are economically disadvantaged."

He pointed to findings that candidates who received at least one offer of admission were less likely to be Pell Grant recipients, were significantly more likely to be free of educational debt or reported significantly lower current education debt, and/or were significantly "more confident in their ability to come up with $2,000 if unexpected need arose within the next month."

"It makes me worried that we are selecting for the aristocrats," Nelson said. "I'm becoming more and more concerned about the economic disparities, more than even the race and gender [disparities]. And that's saying a lot."