Rethinking the Race

Hurdles are by definition barriers. You clear them by leaping, or they will trip you up. Who has the freedom to change the types and placement of hurdles in doctoral programs, and on what basis?

Today, people in many disciplines are considering how time-honored hurdles that students must clear on the way to the Ph.D. may 1) threaten mental health, 2) disproportionately eliminate students from underrepresented groups, particularly students of color and 3) fail to substantively contribute to students’ development. As is often the case, the conventional wisdom about what makes a “good” Ph.D. program may be more convention than wisdom.

Qualifying exams are common to Ph.D. programs. They offer a perfect case study of a traditional hurdle that is up for reconsideration concerning its efficacy for student development and success. We have been studying those exams and the broader transition to candidacy in STEM. It’s part of our collective research agenda to analyze, deconstruct and reconstruct established academic practices -- e.g., admissions, teaching and hiring -- in order to advance racial equity in academe.

We conducted a survey to understand the range of exam structures in physics, chemistry and geosciences; then, we used the results to identify programs that recently changed their exams. We conducted in-depth case studies of two programs’ change processes and the outcomes. The case studies examined the cultural significance of the transition to candidacy and how many faculty members think about altering conventional exams in favor of more developmental, lower-stakes assessments.

Research participants shared stories about rethinking exams as part of broader Ph.D. program improvement efforts. Over and over, they told us that programs “got rid of” the GRE and a traditional qualifying exam because they viewed the costs to equity and well-being as outweighing benefits for selection and learning.

Central to those cost-benefit calculations is evolution in the principles that guide such gatekeeping processes and training. As programs prioritize diversity, equity and inclusion and hold themselves accountable for outcomes and processes that embody those principles, the adequacy of traditional processes comes into question -- especially in their ability to serve students with minoritized racial and gender identities.

The changes that programs imagined, however, depended upon their status in the discipline. In a top-ranked physics program, confidence that their selective admissions process and dissertation prospectus defense were sufficient gatekeeping tools led them to eliminate the high-stakes exam altogether. As one student described to us, "This was already a hard program to get into, and I think the reason they cut the superhard candidacy exam was that they were letting most kids through, but it was just an enormous psychological cost. It was totally unnecessary. Why have two bottlenecks where the second one isn't a good bottleneck, but it's extremely stressful?"

In place of the candidacy exam “bottleneck,” the institution implemented optional diagnostic exams in core content areas for all students upon entering. Those who pass move directly to candidacy and independent research, while others earn candidacy by passing core courses in the first two years of their degree program.

Lower-ranked programs are less selective at the admissions stage, and they may use the exam to signal the quality of the program, their students and their training. When they make change, they may design the new structure to align with field-approved norms. For example, the middle-ranked chemistry program we studied eliminated its conventional sit-down test, but it tied the new candidacy process to producing and defending a dissertation proposal aligned to National Science Foundation or National Institutes of Health proposal guidelines. They then created a scientific writing course where students collectively learn, practice and model the craft of proposal writing. They trained students to compete for grants -- to become players in the academic game.

Our findings are consistent with a U-shaped pattern in organizations’ perceived freedom to innovate based on their positioning in status hierarchies. High-status organizations feel confident to experiment -- to deviate from conventional expectations. Low-status organizations may also feel free to defy conventions because they have little to lose. But middle-ranked organizations seeking to protect or elevate their status tend to conform to field-level norms.

This pattern makes it all the more important that highly ranked Ph.D. programs and resource providers like NSF and NIH use their influence to remove hurdles and set equitable standards for what counts as legitimate practice. They create the ladders that others are trying to climb.

Recommendations

Many people who write about candidacy exams use metaphors to capture their difficulty -- they compared their rite of passage to “an obstacle course and ritual gauntlet,” a “hurdle,” or ambiguous terrain. The need for a more robust gatekeeping tool than the dissertation defense inspired the qualifying exam a century ago, but what is the intensity buying us today -- particularly for students from underrepresented groups who already face additional hurdles in pursuing doctorates? For programs ready to redesign their transition to candidacy, we offer the following recommendations.

• Engage in some introspection. Those of us who direct programs must first reflect on and discuss the underlying assumptions of current models. We should ask questions like “Whom do we imagine when picturing a good or elite scientist?” “Who is excluded from this characterization?” “What concerns do we have about change?” “Who or what are we protecting by not changing?” Only after reckoning with these tensions can we ask and act on the question “What actions will create a more inclusive, expansive understanding of who may grow to become a ‘good’ scientist?”
• (Re)define learning outcomes. The definition of reform is to change by removing faults or abuses. And, indeed, our observation is that Ph.D. programs making change are often more focused on what to get rid of, and why, rather than what they are turning toward. A transformational approach makes space for both. Faculty and doctoral programs should anchor restructuring efforts in specific learning outcomes related to the practice of research and the career trajectories that their students seek. For example, doctoral students may learn how to use disciplinary research to advance the public good. Advisers may actively support such students in relationships or internships with a small number of nonprofit organizations, or even in including a leader from such an organization on their dissertation committee.
• Design for equity. Relatedly, graduate programs today must normalize diverse career paths and validate learners diverse on multiple dimensions and then design equitable structures to support that diversity. Guidelines for universal design for learning offer one framework; liberatory design thinking can also be applied to higher education. As the Association of American Universities Ph.D. Education Initiative argues, many Ph.D. programs are simply not set up to make visible or to value the diversity of strengths that today’s students bring, nor the diversity of graduate students’ professional outcomes.

Policies long defined as requirements for admission, candidacy or degree attainment may not be prerequisites to student success. And depending upon how we define success, those very requirements may actually interfere with it by compromising opportunities, well-being or the sense of belonging of students from already marginalized backgrounds.

Requirements were set at a point in time. As today’s population and the career paths of Ph.D. recipients change, this is a time when requirements deserve reconsideration.

Is your doctoral program revisiting its transition to candidacy or qualifying exam? Share your story with the Equity in Graduate Education Resource Center at bit.ly/GradEdEquity or on Twitter @inclusivegraded.