Diversity beats sameness. We don’t have to search hard for examples: Take Netflix, which makes an airtight case for the power of diversity.
The movie streaming and rental service allows users to rate movies they view, then plugs user data into its proprietary algorithm to generate recommendations. In 2006, Netflix tapped the wisdom of crowds by inviting the public to compete in the Netflix Prize, with the goal of improving the quality of its suggestions for subscribers. The challenge: Increase the accuracy of the Netflix movie-rating algorithm by 10 percent. The reward: $1 million.
Rising to Netflix’s challenge, three researchers assembled a team they called BellKor and created a mathematical model that improved the algorithm by 6.8 percent. BellKor then combined dozens of different models to ratchet up their results even further. And when BellKor joined forces with a competitor, BigChaos, the combination of their models led to still greater improvement. Finally, another team called the Ensemble brought together international members from formerly competing teams, finishing at a dead heat with BellKor and BigChaos. The models submitted by both teams exceeded Netflix’s accuracy by just over 10 percent. And once again, combining the top two models led to even better results.
Scott E. Page (A.B. ’85), an honors math alumnus and current professor of complex systems, political science, and economics, often tells the story of the Netflix Prize to rooms full of listeners to demonstrate the benefits of diversity. The point of the story is simple: When diverse teams exploit diverse resources, they win. But that conclusion didn’t come easily.
The Difference of Difference
Before Page ever set off on what has become a nearly decade-long international lecture tour touting diversity, he’d done a little experiment. He built a few mathematical models to see if he could gain insights about how diverse problem solvers tackle difficult challenges.
“I stumbled on a counterintuitive finding,” he writes about his experiment. “Diverse groups of problem solvers—groups of people with diverse tools—consistently outperformed groups of the best and the brightest. If I formed two groups, one random (and therefore diverse) and one consisting of the best individual performers, the first group almost always did better. In my model, diversity trumped ability.”
The benefits of diversity play out in real life, too—in many different settings. Corporations and industries with diverse teams tend to have greater productivity and profits.
An experiment in Michigan’s Washtenaw County compared mock juries that were either all white or diverse, and it found that diverse juries made fewer errors, discussed more of the evidence, and engaged in longer deliberations.
And there’s growing consensus about Page’s findings. A study coauthored by MIT professor and LSA alumnus Alex Pentland (B.G.S. ’76) affirms Page’s conclusion that groups of expert individuals aren’t guaranteed to succeed. Rather, Pentland and his colleagues found that groups perform better when they’re diverse and empathetic—that is, when they contain more females and if members have the chance to contribute equally.
Page was so bothered and intrigued by the phenomenon that he wrote a book about it. That book, The Difference, moves the conversation about diversity out of the ethical, political, and legal spheres. Instead, it uses math and logic to make the case that higher diversity yields greater tangible benefits. Now, Page gets invited to consult with organizations like the Mellon Foundation, Google, the U.S. Federal Reserve, the U.S. Treasury, NASA, the Society of Women Engineers, and the American Dental Association. They’re taking his argument seriously and running with it in the real world.
Likewise, diverse perspectives are vital to the liberal arts mission of LSA. Departments throughout the College see value in diversity—with benefits both tangible and intangible—and are taking action as a result: They’re developing plans and programs to create a more diverse student body and faculty.
How to Begin
Tim McKay and Vic Strecher (M.P.H. ’80, Ph.D. ’83) knew each other casually from living in the same neighborhood. Though neither of them realized it at the time, they both worked on campus—McKay as a professor of physics and astronomy and the director of LSA’s Honors Program, and Strecher as a professor in the U-M School of Public Health and founder of the Center for Health Communications Research. The first time they met was when they both performed—singing and dancing—in a local community theater company, the Burns Park Players. Only later did McKay find out that Strecher had won an award as the 2010 U-M Distinguished University Innovator of the Year. “I went to hear his talk, because he’s my friend. I had no idea what he did,” admits McKay.
At the award ceremony, Strecher talked about the open-source software package he helped create, which personalizes health advice for patients. “When I heard his talk and learned about his work, it was clear that he had the solution for my big problem,” says McKay, “which was how to personalize the learning experience for students in big classes.”
After Strecher’s talk, the two met to figure out how the public health software could be implemented in McKay’s introductory physics classes. They quickly came up with a plan and raised funds, and within about two years, McKay had created an open-source software package that he calls the Expert Electronic Coaching system, or E2Coach. Whereas before, McKay and other instructors struggled to give students individual attention—particularly in large lecture courses—E2Coach tailors feedback for students according to “diversity in every imaginable dimension,” McKay says.
“Goals, interests, background, demonstrated ability in the class—we’re thinking about the students that we serve in rich and new ways, recognizing that they’re much more diverse than we usually acknowledge, and thinking about how we might serve all of those different groups of people well.” And the data show that students who use E2Coach perform better in class.
McKay continued thinking along these lines, later establishing the Researching Evidence-Based Undergraduate Instructional and Learning Developments project (REBUILD) with faculty from the Departments of Physics, Chemistry, Biology, Math, and Astronomy. REBUILD aims to change the culture of teaching in these departments. The REBUILD faculty want to shift away from conventional teaching methods based simply on habit or tradition. Instead, they want to use scientific methods, data, and evidence to pinpoint the most effective ways of teaching science to undergraduates.
McKay and his colleagues in LSA brought together experts from the School of Public Health, the School of Education, the School of Information, and elsewhere to bolster the REBUILD team—an incredibly collaborative project.
“Michigan created an environment that made these projects much more possible than they would’ve been on any other campus,” McKay says. “The reason we build interdisciplinary projects here is because they bring huge advantages. When you have a team with a major mix of skills, they can do something that no team drawn from just one of those schools can do.”
That includes using reams of data to achieve outcomes that may seem impossible at the start, such as tailoring classes to specially benefit every unique student in the lecture hall.
Making the Difference
Other efforts in LSA aim for more incremental progress toward diversity, though with similarly collaborative teams drawn from campus-wide disciplines. One example: the ADVANCE program, led by Professor of Psychology and Women’s Studies Abigail Stewart, which brings together LSA, the College of Engineering, and the Medical School. The foremost goal of ADVANCE is to foster the most outstanding academic institution possible, according to Carol Fierke, a member of the program’s advisory board, a professor of chemistry, and the dean of the Rackham Graduate School. Originally, ADVANCE worked toward that goal by increasing the number of female faculty in both the science and engineering fields. “We’re missing out on talent,” says Fierke. “People who could be scientists, but aren’t.”
When Fierke arrived at U-M in 1999, only two other female professors ran research labs in LSA’s Department of Chemistry—and one of them retired the year after Fierke settled in. “So our first goal,” she says, “was to increase the number of female faculty.” Later, when the program shifted from national funding to full sponsorship by U-M, ADVANCE included more goals to improve campus diversity. They added underrepresented minorities, then sexual minorities, and now they’re working toward socioeconomic and religious diversity. They’re also promoting diversity in fields beyond science and engineering.
“In the past 15 years, we’ve gone from about 8 percent to about 25 percent female, and 4 percent faculty of color to about 12 percent,” Fierke says, “an enormous change in the face of the [chemistry] department.”
Last spring, ADVANCE invited nearly 50 people to a workshop called NextProf Science. All participants were approaching the faculty phase of their career and had an interest in promoting diversity in universities. With NextProf, Fierke explains, ADVANCE wanted to “show faculty at U-M that diverse candidates are out there. We really need to go out, look for them, and find them.”
Out of ADVANCE sprouted a master’s program that stemmed from a similar realization: If the goal is to foster a more diverse student body, one way to get there is to actively go out and recruit them. The first “bridge-to-Ph.D. program” was established in 2008 by the Department of Ecology and Evolutionary Biology (EEB); they call it the Frontiers Program. The idea spread throughout LSA from there.
Cagliyan Kurdak directs the bridging program in Applied Physics, which they’ve named the Imes-Moore Fellows Program, after two groundbreaking physicists: Elmer Imes (Ph.D. ’18), the second African American in the nation to earn a physics Ph.D., and Willie Hobbs Moore (B.S.E. ’58, M.S.E. ’61, Ph.D. ’72), the first African American woman in the nation to earn a physics Ph.D., both alumni of LSA’s Department of Physics. Kurdak describes the strategy of the program as “bringing in students that will make us proud down the road. We care about students who care about society.”
Kurdak gets excited when he sees that a male applicant is an enthusiastic member of the Society of Women in Physics, or if a student shows entrepreneurial promise by creating products out of salvaged goods as an undergraduate. That type of experience “is not captured in the transcript, not captured in any exam,” he says, but such unique students bring diverse ideas to the program. “We know those students will have a positive impact on society.”
Mark Hunter, a professor in EEB and the former director of Frontiers, elaborates that any of the programs on campus, from the undergraduate to the graduate level, should reflect the society that the University is meant to serve. Fierke agrees, saying, “My guess is that we’re never going to be done.” Because demographics of the general population shift with time, she says, “it’s always going to be a moving target.”
“It’s not like something magic happens at 31 percent that didn’t happen at 28 percent,” says Page about measuring the goals and success of diversity initiatives. “What really matters is allowing people the freedom and opportunity to bring their full selves, their full sets of ideas, their interests, to bear on really hard research problems.”
McKay, for one, wants to do a better job of shepherding students into and through science disciplines, rather than inadvertently weeding out students. He can evaluate the success of E2Coach and REBUILD by keeping track of how many students declare and actually graduate as science majors.
Hunter has kept an eye on what happens to Frontiers master’s students. Do they graduate with a master’s degree, do they continue into doctoral programs, do they find jobs outside of academia? “The expansion of such programs is another metric of success,” he adds, which translates into three or four bridging programs at U-M alone, with more popping up around the country.
Rob Sellers (Ph.D. ’90), a professor and former chair of the Department of Psychology, recently has taken on a new role as vice provost for equity, inclusion, and academic affairs. He plans to look for three major marks of progress in every diversity initiative on campus: Are the programs more diverse than they were? Do we have an environment that everyone feels is inclusive? Do people treat each other with respect?
One metric worth gauging is “the extent to which we become even more excellent at what we do,” Sellers says, “in terms of innovation, problem-solving, and having a greater variety and excellence in our understanding of the human condition.” He’ll lead a campus-wide effort to increase diversity on campus, directly testing Page’s conviction that greater diversity leads to extraordinary outcomes.
Of course, interacting in diverse groups also comes with a challenge—there’s potential for conflict, and communicating across different perspectives and backgrounds simply can be difficult. But, Page says, “You cannot help but confront difference now—at the University and in the world—in a way that you didn’t 20 years ago.” If students prepare now for what they’ll inevitably encounter when they leave campus as alumni, they’ll be equipped to solve problems and innovate in a diverse world—not in spite of diversity, but because of it.
“We’ve got a complex reality. The only way you can possibly prevent really bad outcomes, come up with innovative ideas, predict what’s going to happen, is to have a diversity of ways to think about it,” Page says. Our biggest room for growth is to leverage the power of diversity. “It’s the only way.”