Research studies have consistently identified early academic disparities as the greatest barrier to STEM representation for non-white students. To address the lack of racial and ethnic diversity in STEM, education policies have sought to improve math and science performance in middle and high schools.

A new study refutes these policy assumptions about the STEM pipeline, finding that White English-speaking students are the only group with a positive relationship between early STEM aptitude and STEM enrollment in a higher education institution. For Black and Latine students, early STEM aptitude does not translate into STEM enrollment in the same way.

“Decades of funding to diversify STEM by investing more in early academic preparation may be built on a premise that not only is empirically shaky, but disproportionately favors White students,” says Hua-Yu Sebastian Cherng, the study’s lead author and associate professor of international education at NYU Steinhardt School of Culture, Education, and Human Development.

Cherng and his co-authors analyzed New York City Department of Education data for all public-school students enrolled in 9th grade between 2002 through 2011. They evaluated their records (e.g., math test scores, high school diploma type, school characteristics such as expenditures and student enrollment, and race/ethnicity) from 7th grade through their first semester at the City University of New York (selected for having a large majority of New York City high school graduates compared to any other university in the US). They tracked the outcomes of 10 racial/ethnic groups to find the correlation between early academic success and the declaration of a STEM major at the end of their first year at a two-year or four-year CUNY institution. The 10 groups comprised Asian English speakers, Black English speakers, Latine English Speakers, White English speakers, Spanish speakers, Cantonese speakers, Mandarin speakers, Korean speakers, speakers of languages from South Asia, and Black French speakers.

Their results, published in American Journal of Education, found that disparities in early test scores, high school diploma type, and school characteristics were only correlated to enrollment rates of White English-speaking students majoring in STEM at two-year and four-year institutions.

Additional findings among the racial/ethnic groups include the following:

  • Students speaking Korean, Mandarin, Cantonese, and South Asian languages  (Bengali/Hindi/Punjabi/Urdu), are more likely to declare STEM majors despite not being in the top 25 percent for math scores.
  • For Black English speakers, Black French speakers, Latine English speakers, and Latine Spanish speakers, a far higher proportion show promise in STEM subjects, but do not necessarily declare four-year STEM majors.

“A large body of research has shown the importance of STEM ideation—the idea that young people can see themselves in the field—as being particularly important for groups not represented in STEM,” Cherng says. “To foster these identities, research has pointed to the importance of teachers, families, and community in empowering young people who may not see themselves reflected in STEM careers. I would be remiss if I did not mention the role of discrimination: it is not just that certain groups of young people happen not to get the support they need. Our findings fit into a larger portrait of racial inequality that has been pervasive throughout the history of US education. And the answer to addressing it is not just to think better of others, but to interrogate our fundamental understanding of meritocracy and fallacious explanations of achievement.”

This study was supported by a grant from the National Science Foundation (2017856).