Researchers Track Demographics of Science Talent Search Finalists Over Time
As the 40 finalists of the Intel Science Talent Search descend on Washington this week for a prestigious rite of passage that dates back to 1942, something will be missing. That something is representation from the two racial groups that together comprised one third of public high school graduates in 2010, the most recent year for which U.S. Department of Education data was available. This year, none of the finalists appear to be either African-American or Hispanic.
I say "appear" because the Talent Search does not collect data on ethnicity or race, according to a spokeswoman for the Society for Science & the Public, which runs the contest. So I went through the three years' worth of finalist photos, bios, and names that were available online, a method that is obviously not foolproof. Of the 120 students I viewed, two at most appeared to be Hispanic and none appeared to be African-American. Yet 18 percent of the class of 2010 was Hispanic and 15 percent was black.
This does not mean that the contest discriminates against African-Americans or Hispanics. It is likely that members of these two groups enter the contest at lower rates. However, since racial and ethnic data on participants is unavailable, it is impossible to say for sure.
One thing that is more certain is that the Talent Search's dearth of African-American and Hispanic participants hearkens back to its past. In a study that appears in the current issue of the peer-refereed journal, Teachers College Record, historians Sevan G. Terzian and John L. Rury find that, in its inaugural years of 1942 to 1958, the Talent Search (then sponsored by Westinghouse) set the stage for the popular image of the American scientist as the "brainy white male" who "wears a white coat and works in a lab."
"Success in science was seemingly reserved for certain types of students, while others had to overcome additional obstacles to realize their dreams and aspirations," write Terzian and Rury. "It therefore seems likely that the talent search affected popular conceptions of 'meritocracy' and discouraged many African-American and other minority students from pursuing scientific careers."
Terzian and Rury report that an analysis of 220 of the 1942 finalists and semifinalists found just one African-American. There is no mention of Hispanics. Additionally, an unrelated 2006 study of four historical cohorts found similar results for 160 students who were finalists between 1965 and 1985. That study, which appeared in the peer-refereed Journal of Adult Development, was conducted by Gregory J. Feist, an associate professor of psychology at San Jose State. Feist found that just 3 percent of finalists were Hispanic. None were African-American.
Asked about historic demographic profiles, a spokeswoman for the Society for Science & the Public responded that the Talent Search "is now, as it has always been, a merit-based program" that is "open to any U.S. high school senior." She added: "We are continually looking for ways to reach out to new audiences and encourage their interest and participation; not only in the Intel STS and our other educational competitions, but also in any form of hands-on independent research, as well as staying informed and aware of the latest scientific news and discoveries."
Though it is unclear whether this outreach played a role, the demographics of contest participants have undergone major changes since 1942. For example, there were no Asian-American finalists or semifinalists in 1942. By my count, about two-thirds of the 2012, 2013, and 2014 finalists were of Asian descent. That compares to just over 5 percent of the public high school class of 2010. Feist also found that Asian-Americans were overrepresented in the 1965-1995 cohorts. One explanation is Feist's finding (confirmed by other researchers) is that, throughout the contest's history, recent immigrants have been disproportionately represented among winners. In the cohorts Feist studied, 36 percent of finalists had a foreign-born or first-generation father and 40 percent had a foreign-born or first-generation mother.
Female representation has also increased. In the first 16 years of the contest, 23 percent of the finalists were female, a ratio that was supposed to reflect the rate at which high school girls applied to the Search. This approach may actually have skewed the results if females in a particular year happened to be particularly strong (and thus deserve overrepresentation among winners) or weak (and thus deserve underrepresentation). That quota no longer exists. And the Talent Search does track gender, so I can say with confidence that, in the past three years, as well is in the past decade, the share of female finalists has increased substantially since the 40s and 50s. It is now 42 percent.
Despite this increase, females remain underrepresented because they comprised 50 percent of the public high school class of 2010.
As in the 40s and the 50s, certain areas of the country continue to be overrepresented among Talent Search winners. Although New York housed 9 percent of the nation's high school seniors back then, the state was home to 33 percent of 1942-1958 finalists.
New York's dominance has persisted. In the past decade, 26 percent of finalists have hailed from that state, which was home to 6 percent of the nation's 2009 high school graduates. As in the 1940s and 50s, no other state sent more finalists to the Talent Search.
In their study, Terzian and Rury found that winners were significantly more likely to come from states that spent more money on education. In the 1940s and 50s, New York spent more money per pupil than any other state. Today, the state remains one of the nation's top spenders on education.
Two finalists from the contest's early years said that their schools' resources helped them develop their interest in science.1947 finalist Leon Cooper, 84, used the facilities at New York City's selective public Bronx High School of Science to create a more penicillin-resistant strain of bacteria.
"I had a laboratory at Bronx Science," said Cooper, a Brown University professor who was awarded a Nobel Prize for physics in 1972. "I could hardly have done that at home. It was something I used to do after school. I spent my afternoons in the laboratories. It was wonderful. It was a wonderful experience."
Similarly, 1955 finalist Roald Hoffmann, 76, recalled that the science facilities were "excellent" at his alma mater Stuyvesant, another selective public high school in New York City.
"Stuyvesant was the greatest collection of intellects I had ever known in my life," recalled Hoffmann, a Cornell University emeritus professor who won a Nobel Prize for chemistry in 1981. "The teachers were wonderful. They were products of the Depression. Some had PhDs but teaching was a profession to enter in those days. They provided a lot of incentive and support for any student who wanted to move ahead in science, toward research. The teachers talked this up."
Hoffmann also availed himself of the city's nonschool resources. After reading a Scientific American article about cosmic rays, he was able to find a local professor who was doing research on that topic. His work with that professor's data became his prize-winning Westinghouse project.
An overarching question is whether the contest results, (and, by extension, their demographics), really matter. When it comes to the impact on the lives of the individual contestants, the jury is still out. Feist did find that the contest was relatively good at identifying students who would go on to pursue education and jobs in science. More than 80 percent of the finalists he studied and 70 percent of finalists overall earned PhDs or MDs, as compared to about 1 percent of the general population . Nearly 60 percent were conducting scientific research in the realms of academia, industry or government. Additionally, Feist notes that, among the 2,080 students who were selected as finalists between 1942 and 1994, five earned Nobel prizes, and 28 became members of the prestigious National Academy of Sciences. By comparison, other programs have had more limited success at identifying young people who go on to successful careers in science and/or math. For instance, just a quarter of participants in Stanley's Study of Mathematically Precocious Youth went on to earn doctorates.
However, these analyses, like others I found, all suffered from the classical chicken-versus-egg question: Did the Talent Search merely identify students who were already destined to become scientists? Or does finalist status in itself shape and contribute to the winners' scientific careers?
My highly unscientific sample of three early finalists who won Nobel prizes suggests that the contest may have had an impact, albeit a limited one. None of the three said the projects they submitted formed the basis for important, subsequent work. But Cooper believes his finalist status might have helped him get admitted to Columbia University as an undergraduate. A finalist from 1950, Shelly Glashow, 81, who was awarded the Nobel Prize for physics in 1979, said the contest boosted his self-confidence and introduced him to new friends who shared his award trip to Washington, D.C. As for Hoffmann, he said the money he won helped him buy books and other necessities for college. He said he was sure that his finalist status helped him obtain the first of a series of summer jobs that kept him interested in research during his early undergraduate years, when he found his formal science coursework to be less than inspiring.
"Without those jobs, I would be a good doctor in Brooklyn and my mother would have been happier," he quipped.
Although the impact on individual contestants' lives is difficult to assess, Terzian and Rury found that, in the early years they studied, the Talent Search as an institution was important in that it both reflected and contributed to perceptions, both accurate and inaccurate, of American scientific endeavors.
"Ultimately," they conclude, "this national competition reflected social and cultural forces that shaped the science professions in a crucial period of their growth, and may have represented a lost opportunity to make scientific training more truly meritocratic at a formative time in its development."