Bias Isn't Just A Police Problem, It's A Preschool Problem

Sep 28, 2016
Originally published on September 30, 2016 8:18 am

First, a story:

Late one night, a man searches for something in a parking lot. On his hands and knees, he crawls around a bright circle of light created by a streetlamp overhead.

A woman passes, stops, takes in the scene.

"What are you looking for? Can I help?"

"My car keys. Any chance you've seen them?"

"You dropped them right around here?"

"Oh, no. I dropped them way over there," he says, gesturing vaguely to some faraway spot on the other side of the lot.

"Then why are you looking here?"

The man pauses to consider the question.

"Because this is where the light is."

New research from the Yale Child Study Center suggests that many preschool teachers look for disruptive behavior in much the same way: in just one place, waiting for it to appear.

The problem with this strategy (besides it being inefficient), is that, because of implicit bias, teachers are spending too much time watching black boys and expecting the worst.

The study

Lead researcher Walter Gilliam knew that to get an accurate measure of implicit bias among preschool teachers, he couldn't be fully transparent with his subjects about what, exactly, he was trying to study.

Implicit biases are just that — subtle, often subconscious stereotypes that guide our expectations and interactions with people.

"We all have them," Gilliam says. "Implicit biases are a natural process by which we take information, and we judge people on the basis of generalizations regarding that information. We all do it."

Even the most well-meaning teacher can harbor deep-seated biases, whether she knows it or not. So Gilliam and his team devised a remarkable — and remarkably deceptive — experiment.

At a big, annual conference for pre-K teachers, Gilliam and his team recruited 135 educators to watch a few short videos. Here's what they told them:

We are interested in learning about how teachers detect challenging
behavior in the classroom. Sometimes this involves seeing behavior before it becomes problematic. The video segments you are about to view are of preschoolers engaging in various activities. Some clips may or may not contain challenging behaviors. Your job is to press the enter key on the external keypad every time you see a behavior that could become a potential challenge.

Each video included four children: a black boy and girl and a white boy and girl.

Here's the deception: There was no challenging behavior.

While the teachers watched, eye-scan technology measured the trajectory of their gaze. Gilliam wanted to know: When teachers expected bad behavior, who did they watch?

"What we found was exactly what we expected based on the rates at which children are expelled from preschool programs," Gilliam says. "Teachers looked more at the black children than the white children, and they looked specifically more at the African-American boy."

Indeed, according to recent data from the U.S. Department of Education, black children are 3.6 times more likely to be suspended from preschool than white children. Put another way, black children account for roughly 19 percent of all preschoolers, but nearly half of preschoolers who get suspended.

One reason that number is so high, Gilliam suggests, is that teachers spend more time focused on their black students, expecting bad behavior. "If you look for something in one place, that's the only place you can typically find it."

The Yale team also asked subjects to identify the child they felt required the most attention. Forty-two percent identified the black boy, 34 percent identified the white boy, while 13 percent and 10 percent identified the white and black girls respectively.

The vignette

The Yale study had two parts. And, as compelling as the eye-scan results were, Gilliam's most surprising takeaway came later.

He gave teachers a one-paragraph vignette to read, describing a child disrupting a class; there's hitting, scratching, even toy-throwing. The child in the vignette was randomly assigned what researchers considered a stereotypical name (DeShawn, Latoya, Jake, Emily), and subjects were asked to rate the severity of the behavior on a scale of one to five.

White teachers consistently held black students to a lower standard, rating their behavior as less severe than the same behavior of white students.

Gilliam says this tracks with previous research around how people may shift standards and expectations of others based on stereotypes and implicit bias. In other words, if white teachers believe that black boys are more likely to behave badly, they may be less surprised by that behavior and rate it less severely.

Black teachers, on the other hand, did the opposite, holding black students to a higher standard and rating their behavior as consistently more severe than that of white students.

Here's another key finding: Some teachers were also given information about the disruptive child's home life, to see if it made them more empathetic:

[CHILD] lives with his/her mother, his/her 8- and 6-year-old sisters,
and his/her 10-month-old baby brother. His/her home life is turbulent, between having a father who has never been a constant figure in his/her life, and a mother who struggles with depression but doesn't have the resources available to seek help. During the rare times when his/her parents are together, loud and sometimes violent disputes occur between them. In order to make ends meet, [CHILD's] mother has taken on three different jobs, and is in a constant state of exhaustion. [CHILD] and his/her siblings are left in the care of available relatives and neighbors while their mother is at work.

Guess what happened.

Teachers who received this background did react more empathetically, lowering their rating of a behavior's severity — but only if the teacher and student were of the same race.

As for white teachers rating black students or black teachers rating white students?

"If the race of the teacher and the child were different and [the teacher] received this background information, severity rates skyrocketed," Gilliam says. "And the teachers ended up feeling that the behavioral problems were hopeless and that very little could be done to actually improve the situation."

This result is consistent with previous research on empathy, Gilliam says. "When people feel some kind of shared connection to folks, when they hear more about their misfortunes, they feel more empathic to them. But if they feel that they are different from each other ... it may actually cause them to perceive that person in a more negative light."

It's impossible to separate these findings from today's broader, cultural context — of disproportionately high suspension rates for black boys and young men throughout the school years, of America's school-to-prison pipeline, and, most immediately, of the drumbeat of stories about black men being killed by police.

If implicit bias can play a role on our preschool reading rugs and in our classrooms' cozy corners, it no doubt haunts every corner of our society.

Biases are natural, as Gilliam says, but they must also be reckoned with.

The good news, if there is such a thing from work such as this, is that Gilliam and his team were ethically obligated to follow up with every one of the 135 teachers who participated in the study, to come clean about the deception.

Gilliam even gave them an out, letting them withdraw their data — for many of them, the lasting proof of their bias.

Only one did.

Copyright 2018 NPR. To see more, visit


Racial bias has been much in the news in the wake of a series of police shootings of unarmed African-American men.

During Monday night's presidential debate, Hillary Clinton said it's not just a problem for police.


HILLARY CLINTON: I think, unfortunately, too many of us in our great country jump to conclusions about each other. And therefore, I think we need all of us to be asking hard questions.

MONTAGNE: That would include those who help shape the youngest among us. A new report from the Yale Child Study Center has found evidence of implicit racial bias among preschool teachers.

Here to tell us more about those findings is Cory Turner of the NPR Ed Team.

Good morning.

CORY TURNER, BYLINE: Good morning, Renee.

MONTAGNE: Sketch out for us how this study was organized, which, by the way, I imagine bias is a very hard thing to study and to nail down.

TURNER: It's very hard. Because the moment you tell your subject that you're studying bias, you're probably not going to get an honest answer. So it's what makes this study so interesting because the team from Yale had to create something essentially tricky.

They recruited roughly 135 preschool teachers and had them watch video footage of four kids - a black boy, black girl, a white boy and a white girl. And they were told watch the footage. Your job is to press the enter key on your keyboard every time you see a behavior that could become a potential challenge.

Now here's the tricky part. While the teachers were doing this, the researchers were using eye scan technology to measure the trajectory of their gazes to see which kids they were watching, anticipating bad behavior.

MONTAGNE: And what happened?

TURNER: Well, they spent the most time watching the black boy. And interestingly, this is true of both white and black teachers.

The researchers then asked the teachers explicitly, which child did you feel required the most attention? Forty-two percent said the black boy, 34 percent the white boy, just 13 percent the white girl and 10 percent the black girl.

MONTAGNE: So teachers were not necessarily seeing bad behavior where there wasn't any, but they spent the most time watching the black boy, as you've just said, expecting bad behavior.

TURNER: Exactly. In fact, in the video, there was no bad behavior at all.

And when I spoke with lead researcher Walter Gilliam, he told me, you know, the problem with this is if you look for something in only one place, well, that's the only place you can typically find it.

And that helps make sense of what we already know about preschool suspensions, which is that black children make up about 19 percent of preschoolers in the U.S., but make up nearly half of preschoolers who get suspended.

MONTAGNE: And, Cory, this study was in two parts. Walk us through the second half of it.

TURNER: This is where it gets even more interesting.

So teachers were each given a paragraph to read that described a hypothetical student - sometimes a boy, sometimes a girl - doing some pretty disruptive things in the classroom, hitting, pushing, throwing toys. And the student was randomly assigned a stereotypical name. And the teachers were asked to rate the severity of these bad behaviors.

But where it gets really interesting is when some of these teachers were also given background information about the child, saying that this child comes from a relatively poor household, lives mostly with his or her mother. Father has been in and out, mother is depressed, works three jobs. The researchers really wanted to find out would this make the teachers more empathetic to the kids?

And it did when the teachers were of the same race as the student. But interestingly, it had the opposite effect for teachers who were of a different race from the student. It actually made them less empathetic.


TURNER: You know, Renee, that's just a really complicated question that this research doesn't entirely answer.

I talked to lead researcher Walter Gilliam, and he did tell me this.

WALTER GILLIAM: Teachers ended up feeling that the behavioral problems were hopeless, and that very little could be done to actually improve the situation.

TURNER: Renee, I will say there is one glimmer of hope here. Researchers all agree that if we're really going to confront the challenges of implicit bias in our schools and everywhere else, we have to talk about it. We have to be honest about it.

And Gilliam told me that because there was deception involved in his study, he was ethically obligated to go back to these 130-some-odd teachers that he worked with, and to tell them I deceived you. You have the right to withdraw your data from my study. And he said he was pleasantly surprised that only one did.

MONTAGNE: That's Cory Turner of NPR's Ed team. Thanks very much.

TURNER: Thank you, Renee. Transcript provided by NPR, Copyright NPR.