November 2012

Dimensions of diversity

We are all shaped by our life experiences. I certainly was influenced strongly by the community in which I grew up. My father was a math professor at Stanford University, and my mother was a physician who worked as a researcher at Stanford Medical School during my early childhood. We lived in a house on the university campus, and I attended an elementary school whose student population was about equally divided between children of university faculty and staff members, children of graduate students and children from the surrounding town. On one hand, the school was marvelously diverse, primarily by virtue of the children of graduate students; among about 400 students (grades K–6), between 15 and 20 countries were represented at the annual international day. On the other hand, the school was quite homogeneous. For the majority of students, one or both parents were professionals with advanced degrees. Indeed, it wasn’t until I was 12 and started junior high school that I truly realized that the question, “What does your father do?” did not mean, “In what field is your father’s master’s degree or doctorate?” Most importantly, I had little insight at the time that the world I was experiencing was not the same as that experienced by all other people.

More than 30 years later, I became responsible for a number of programs intended to increase the diversity of the scientific workforce as a component of my position as the director of the National Institute of General Medical Sciences. While I was at the NIGMS, we examined the goals and structure of these programs. Perhaps the most interesting and important discussions that I participated in with National Institutes of Health staff and scientists throughout the United States involved questions of great importance: Why does diversity in the American scientific workforce matter? What are the benefits of a diverse workforce to individuals, to institutions, to the research enterprise, to the nation? Many of the discussion points focused on two major observations that became increasingly apparent during our meetings.

The first observation is that groups with members from highly diverse backgrounds simply perform differently (and usually more effectively) than groups with little diversity. This has been observed empirically in many settings and has been studied more formally as described in Scott Page’s book “The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies” (1). This observation is certainly consistent with many trends in modern science, such as the interest in interdisciplinary research. Interdisciplinary research teams often have the power to solve hard and important problems because each member of the team brings (at least in principle) a perspective and skill set to the problem that some or all other members of the group may lack.

What does this have to do with dimensions of diversity such as race, ethnicity and sex? At the population level, the reality in our society is that our experiences do depend on these factors. For example, I had a great opportunity to hear from students directly when I attended the 10th anniversary of the Annual Biomedical Research Conference for Minority Students in 2010 (2). This conference attracted more than 2,000 (primarily undergraduate) students from around the country, many of them from black, Hispanic or Native American backgrounds. I learned firsthand how passionate many of these students were about science, despite — or perhaps because of — the fact that they came from backgrounds where interest in a career in research was quite rare.

At this conference, I also presented a keynote lecture. Initially, I was planning to give a talk about the NIH and the programs that the NIH runs to help students — particularly those from groups that have been traditionally underrepresented in science — initiate and progress through careers in biomedical research. However, as I prepared, I wondered how well the audience would connect with what I worried would be a hackneyed talk about bureaucracy. My wife recently had introduced me to “The Last Lecture” (3). This powerful book was written by Randy Pausch, a leading computing science professor at Carnegie Mellon University who was diagnosed with terminal pancreatic cancer and gave a lecture at CMU entitled “Achieving Your Childhood Dreams.” In his lecture (4), which went viral on YouTube (more than 15 million views to date), Pausch described his childhood ambitions and how he achieved many of them through determination, powerful mentorship at key points in his life, and support from his friends and family. I was tremendously moved by the book and video, and I decided to change course, shaping my lecture around the same theme: describing my dreams, how I had been able to achieve some of them (as well as some things that I had never imagined possible), and how there remained other dreams to pursue. Though I enjoyed giving the lecture itself, by far the most gratifying part of the experience was the conversations that I had with a number of seminar participants after the talk. One student, in particular, made comments that still resonate with me today, saying, “A lot of speakers at these meetings talk to us about being minorities. News flash: We know that we are minorities, and we don’t really think you have much insight to offer us about that. But your lecture got me thinking about what I really want to achieve in my life.”

The diverse perspectives I experienced at the conference changed my views on a range of topics, and I hope that I also was able to influence other people’s thinking.

The second major observation is that it is essential that there be diverse role models for the scientists of tomorrow. Many of us have been influenced strongly by particular teachers or other individuals over the courses of our lives. Part of this impact comes from our ability to identify with those individuals and to see ourselves in their shoes. Indeed, the importance of role models is borne out by empirical studies. For example, a recent study (5) examined the performance of male and female students at the Air Force Academy based on the sex of their teachers in 2009. The Air Force Academy was used for this study because the curriculum includes a number of mandatory courses and students are assigned to particular professors randomly, simplifying the analysis considerably. The study revealed a correlation between female students’ math performance in introductory courses and having female faculty; there was no such correlation noted among male students. Furthermore, having female professors for introductory classes increased the likelihood that female students would pursue additional math and science courses later in college.

Unfortunately, although considerable progress has been made with regard to sex, the biomedical workforce is not yet representative of American society with regard to race and ethnicity. This is particularly true for university faculty. This has implications with regard to the consequences of implicit biases that we all have but of which we are frequently unaware. One can hypothesize that increasing representational balance among faculty could be a key strategy for increasing representation in the broader workforce.

Two other recent studies reveal some additional challenges related to diversity. The first is a study showing that simply changing the name on a résumé from “John” to “Jennifer” on applications for a laboratory manager position resulted in a statistically significant decrease in the rating of the applicant as competent and hirable (6). Previous studies have shown similar influences of indications of racial or ethnic backgrounds in other employment settings. These results presumably reflect the impact of implicit biases that affect our judgments even when we are not aware of them.

The second study is an NIH study that revealed differences in funding success rates among different racial and ethnic groups (7). In particular, the success rates for applications from black scientists were statistically significantly lower, even after correcting for other factors, such as institution and publication history. There are several nonmutually exclusive explanations for these observations, including review bias, either implicit or explicit, and the influence of a range of factors on the characteristics of the applications. These studies highlight some of the challenges we face in the 21st century in trying to achieve representational balance in the scientific community.

Moving forward, I hope that students, faculty and institutions alike can refocus our energies on what we want to achieve together while valuing individual differences. To that end, it will help if we all become more aware of our own biases, just as I learned many years ago that not everyone grows up on a college campus. For a thought-provoking experience, I highly recommend exploring some of your own biases on tests available online (8). Diversity enhances the richness of the fabric of the scientific enterprise, including the questions that are asked and how they are approached. We will all learn from others with different perspectives and skill sets and will be able to contribute our own perspectives to benefit others.

  1. (1) Page, S. The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Society, Princeton University Press (2007).
  2. (2) 
  3. (3) Pausch, R. and Zaslow, J. The Last Lecture, Hyperion Books (2008).
  4. (4) 
  5. (5) 
  6. (6) Moss-Racusin, C. A. et al. Proc. Natl Acad. Sci. USA Published ahead of print Sept. 17, 2012.
  7. (7) Ginther, D. K. et al. Science 333, 1015–1019 (2011).
  8. (8) 


Photo of Jeremy BergJeremy Berg ( is the associate senior vice-chancellor for science strategy and planning in the health sciences and a professor in the computational and systems biology department at the University of Pittsburgh.

Randy Pausch’s last lecture at Carnegie Mellon University

found= true2051