April 9 — 11:44 am, 2019

Princeton report shows correlation between higher black student discipline and implicit bias

Researchers combined federal Department of Education records with data collected from 1.6 million visitors to the Project Implicit website to reach their conclusions.

Naomi Elegant

In public schools across the country, black students experience higher rates of disciplinary action compared to their white classmates. According to a new report from Princeton University researchers, this disparity is greater in counties that demonstrate higher levels of racial bias as shown by an ongoing data-gathering effort called Project Implicit.

The report’s authors, Stacey Sinclair and Travis Riddle, combined federal data from the Department of Education with data from Project Implicit. The project is run by a nonprofit organization that collects information about people’s biases by having them take tests online. The researchers analyzed more than 3,100 counties in the United States and, using a statistical model, they found a correlation between higher recorded levels of racial bias and higher disciplinary disparities between black and white students.

“Our research joins a wealth of other findings suggesting that racial bias is contributing to disciplinary disparities,” said Sinclair, a professor of psychology and public affairs at Princeton. “In terms of a policy fix, it’s hard to say what that would be. Our particular research identifies a relationship between racial bias and disciplinary disparities, but does not specify the reasons for this relationship.”

Project Implicit, founded in 1998 by scientists from the University of Washington, Harvard University, and the University of Virginia, “investigates thoughts and feelings that are largely outside of active awareness or control” through online tests and questionnaires, according to its website. Participation is self-selected and voluntary.

“We show that county-level estimates of racial bias, as measured using data from approximately 1.6 million visitors to the Project Implicit website, are associated with racial disciplinary disparities across approximately 96,000 schools in the United States, covering around 32 million white and black students,” the authors write in an abstract of their study, which was published April 2 in the journal called Proceedings of the National Academy of Sciences.

Riddle, a postdoctoral researcher in Princeton’s department of psychology, said that finding a causal relationship is “an important next step,” especially for determining possible methods to counteract the behavior and reduce these disparities.

“One of the challenges is just that we don’t really know which thing should be intervened on,” Riddle said. “And this is compounded by the fact that intervening on these kinds of things can be pretty tough.”

The research focused on racial bias in five common disciplinary actions — in-school arrests, expulsions, law enforcement referrals, and in-school and out-of-school suspensions. For all five, black students had higher rates than whites. For suspensions, the rate was nearly four times higher, with 13.46 percent of black students receiving out-of-school suspensions vs. 3.5 percent of white students.

Riddle said that other studies have shown that such disciplinary action can be damaging to students — in-school arrests, for example, have been associated with an increased risk of antisocial behavior and dropping out of school. They can also prove to be the first step of the school-to-prison pipeline, in which children are, in effect, funneled out of school and into the criminal justice system.

To make sure these results were specific to racial biases, the researchers performed the same analysis using sexuality bias and found no meaningful correlation between higher rates of disciplinary action and sexuality bias. A separate report published in 2018 found that disciplinary disparities between black and white students also existed regardless of school poverty levels and type of public school.

Riddle and Sinclair have been working on the study since late 2017.

the notebook

Our news is free to read, but not to report.

support local journalism