Merit, Diversity and Grad Admissions

Litigation and political battles about affirmative action tend to focus on undergraduate or professional school admissions, which are supervised by admissions professionals. In Ph.D. admissions, faculty members are the key players. And although they too must weigh the relative value of various measures of merit, and how much diversity should be considered a form of merit, a separate qualification or not considered at all. A new study aims to lift the veil on this type of admissions -- at least at a group of top programs.

Among the key findings: Initial reviews of candidates tend to be dominated by traditional definitions of merit (grades and test scores), with little if any attention given to issues of diversity. When diversity comes into play later in the process, professors on admissions committees seem to have a range of motives -- some reflecting more than others a commitment to diversifying their fields. For instance, qualifying for more institutional funds for graduate fellowships seems a strong motivator for departments to give additional consideration to minority candidates.

The study, by Julie R. Posselt, assistant professor of education at the University of Michigan, has just been published in American Journal of Education (University of Chicago Press), and an abstract is available here. For her study, Posselt observed admissions committee meetings and interviewed participants at 10 highly selective graduate programs, all highly ranked in their disciplines. The programs are at three research universities and across a wide range of disciplines. Posselt writes that her agreement with institutional review boards required that she grant the professors and their institutions complete anonymity. (She does note the disciplines of various professors.)

Posselt notes that there were of course lots of differences from faculty member to faculty member on what they most valued in graduate school candidates. But she writes that her emphasis in the paper is on themes that were present with many or most departments and faculty members.

For the Ph.D. programs Posselt observed, which don't lack for applicants, the clear trend was that you only make the first cut with high test scores and top grades, preferably from prestigious undergraduate institutions. While many faculty members care very much about diversity issues, she writes, that doesn't factor in when the winnowing starts.

"Across programs the initial conceptualization of merit rested on numerical indicators of conventional academic achievement that can be used to quickly compare students," she writes. "They included GRE and TOEFL scores and grade-point averages, which faculty often contextualized by institutional prestige and curricular rigor. An astrophysicist admitted, 'I would say -- and you will see it in our discussions -- it’s very unlikely that we would consider anyone who has a low subject GRE.' One of his colleagues concurred: 'If you don’t score high, you’re probably not going to make the cut.' "

Posselt said that many faculty members justified their reliance on traditional measures of merit by talking about why they are "risk averse" in admissions decisions. A classicist in one of the departments she studied told her that "there’s always a tension here because we’re always under pressure to have good numbers for completing a program, completing it in a reasonable amount of time, and so on. The effect of that is to make you risk averse because it’s not that hard just to go for the students you’re pretty confident can get through.”

A physicist told her: "If you work with the student so closely and then he walks away or doesn’t make it, then it’s a waste of his time and in a way, I mean, it’s our mission to teach, but I’d rather spend my time teaching somebody who actually can continue my mission and then teach other students than somebody who realizes, 'It’s just too difficult. I can’t do it.' "

Grades and test scores are also highly valued by the minority faculty members who were interviewed, even those who said that they planned to make a push on diversity issues later in the process. One minority sociologist is quoted as saying: "You have to be above a bar, and then we can ask the diversity question."

Posselt also found that faculty members frequently were convinced that an emphasis on high grades and test scores protected them. Several used the word "spooked" to describe reading the application of someone who reminded them of an unsuccessful past student.

Looking for Diversity

When faculty members are looking at those who have made a first or second cut, diversity appears to be more of a factor, Posselt writes, with different professors having different frameworks for considering the issue. Many spoke of an "obligation" to improve their discipline by making it more diverse. Sometimes this related to specific issues within a discipline. For example, philosophy professors -- aware that their field currently has a reputation for being "macho" (their word) -- said that they wanted to admit more women.

Faculty members also viewed themselves as competing for minority talent with other top programs and assumed that they could predict who would be attractive to other departments as well. Many professors said that they view a healthy minority graduate enrollment as an important sign of a department's prominence.

Because the faculty members value a high yield (the percentage of admitted applicants who accept admissions offers), they talk quite a bit about whether minority applicants are likely to enroll, and those comments could help or hurt minority applicants. "Of an underrepresented minority student who attended an Ivy League institution, a philosophy professor made the offhanded comment, 'Sounds like he’ll get in everywhere. Everyone will love him,' " Posselt writes.

She describes these as comments heard in committee discussions about minority applicants: “Can we get her?” “Who are we going to get? It’s a gamble,” and “We’ll lose him to Princeton and Caltech.”

Competition is particularly intense, professors said, for black students. And many professors talk about the competition issue along with their goal of getting new and different ideas into their programs. One economist is quoted as saying: "I think people care most about intellectual diversity -- that people arrive with different interests, different preparations, and are likely to write very different theses on different topics. I think that’s the type of diversity we would value most. I mean gender is an issue in that we get good -- we get top-notch women as well top-notch men. Black -- we get fewer blacks. It’s true. But we do try -- in the past we’ve tried to attract them. But then they get the same attractive offers from Columbia and Yale and Stanford and Berkeley and so forth. So it’s a small group typically who get a lot of attention."

A discussion of candidates in an astrophysics department showed that talking about one's status as an applicant who is a woman in a predominantly male field draws different responses from admissions committee members. A female applicant being discussed at one of the observed committee meetings had written in her personal statement about her experiences in science, including sometimes being made fun of by teachers and fellow students. One committee member asked, "Is it enough to be a woman in science?" A male faculty member defended the teacher who had made fun of the student, and asked if she might arrive "with an axe to grind."

The woman was admitted when another committee member (also male) defended her by noting that she wasn't just complaining, but was doing things at her undergraduate institution -- such as organizing a lecture series -- that promoted change.

"This example demonstrated the tendency for identities and experiences of so-called diversity candidates to be scrutinized at a level that applicants from majority backgrounds were not," writes Posselt.

Even as faculty members on committees expressed philosophical commitment to diversity, Posselt observed financial motivations at play. Some universities offer extra funds for minority graduate students, so that a fellowship might be paid for from general university funds and not departmental funds. Where such incentives exist, they appear to have a strong impact, Posselt writes.

She quotes one sociology professor, who had this answer to how his department thinks about diversity: "I think roughly in whatever way the university will pay for. Our conceptualization is the university’s conceptualization, and that’s putting it a little harsh. But because the university’s commitment is quite good and there are lots of incentives, we don’t need to add any interests that are beyond what are well established through straightforward incentives. So we’ll just do whatever. We define diversity as the university defines it. If this person’s going to be eligible for some sort of special resources, then we’re all for it. I don’t see much for it beyond the university’s commitment as implemented through a variety of programs that give us resources we need to be more diverse."

Posselt concludes her paper by questioning the current two-step approach in which diversity doesn't get much attention at all initially, but then is added as a topic for consideration. She writes that "many of the students whose diversity contributions might have been considered assets had already been filtered from the pool."

Departments may make only limited diversity progress in such an approach, she cautions. "The current two-tier review process relegates diversity to a secondary consideration, and it makes a standard of inclusive excellence conditional on conventional achievement," she writes. "None of the 10 programs began with diversity as a major criterion. Yet, if a program deeply values diversity, then when they value it matters for the outcomes they are likely to achieve."

In an email interview, she said that she took care in the order of her interview questions so that people would answer honestly. "I was concerned that people would report more socially desirable comments about diversity when talking to a researcher," she said. "One way I tried to account for this was to save questions about diversity for the end of the interviews, to allow some time for trust and rapport to develop. This concern was also big part of why I wanted to observe meetings that produced official admissions outcomes for the department. Even then, though, I knew there was a risk that committee members would change their behavior (on diversity of other matters) under observation, so I took it as a good sign for the validity of the data when people would say inflammatory, rude, even illegal things — as often they did."