Like most Americans, Andrew Meltzoff, co-director of the University of Washington Institute for Learning & Brain Sciences, had heard that girls can't do math. Research shows that in the U.S., both adults and older children believe that math is a male activity.
But Meltzoff wanted to know how early children "catch" stereotypes like this, and when they learn social biases, more generally. He found that stereotypes are ingrained by second grade – before children even learn their multiplication tables.
He sat down with KPCC's early childhood reporter Devin Browne at the Simms/Mann Institute Think Tank to explain more.
You're most known for your research about infant imitation and early cognition. How did you get started studying stereotypes and little kids?
It's connected to the idea that even from birth and early infancy, before there’s language, children are highly socially attuned — they watch what we do and they pick up lessons from watching us.
So at a more general level, I became interested in whether they’re picking up our stereotypes, whether they’re picking up the idea that girls are not supposed to do math, and if so, would that affect their learning in schools? So, we have done a series of studies now to look at how early on children pick up stereotypes and made a discovery that as early as second grade, about 75 percent of the little girls and boys in Seattle, according to our experiment, have caught the stereotype that girls don’t do math.
And what’s so shocking about this, is that in school, we’re really teaching multiplication tables usually by third grade, so here in second grade – before the children have even learned their multiplication tables – little girls have caught the stereotype: math is not for me. Math is not for girls.
And we think this is something very important to discover and something that we can help change what societal stereotypes are about or to try to help the little girls and little boys to try to resist or overcome those stereotypes.
How do you know that so many little girls feel this way?
We devised a special test where [children] could push a button about what object or visual image went with boy and what one went with girl. They were supposed to push, for instance, a left button if this object is something that went with girls, or a right button if it was something that went with boys. It’s sometimes called an implicit stereotype or implicit bias test, and we were able to adapt this test so it could be used for little children, elementary school children.
So when we showed children images of for instance, numbers, they would press the boy button. They did that spontaneously; they thought numbers went with boys. When we showed them letters, they would push the girl button. We were able to discover with these implicit tests that kids as early as second grade shared some of the stereotypes that [American adults] had.
Does that mean that in preschool and kindergarten kids still math as something gender neutral?
That’s a terrific question. We have found out that by second grade kids are highly stereotyped with these academic subjects. We think that there’s an onset of that and we’re not sure where that onset is, but yes, somewhere in kindergarten or first grade, it must start. But we’re able to find significant effects by second grade.
One of the things that we’re quite interested in is the development of these stereotypes: How early on do American kids catch the stereotype? What’s the role of mother/father stereotypes or even their school teachers’ stereotypes in conveying that to the children themselves? So the intergenerational transfer of stereotypes is a very, very important thing. It’s sort of like super-charged imitation. The infants are imitating the actions we do and the elementary school kids are thinking something like the following: I’m a girl — that’s gender identity — I’m a girl and I look out into the world and see that society acts as though girls don’t do math. That’s a stereotype, girls don’t do math. They pick up that stereotype and then they begin to apply it to themselves and think, “I’m a girl, girls don’t do math, therefore, I don’t do math.” And it begins to affect what we call their self-concept.
So there’s a chain of development there that we’re uncovering, that first they understand what gender they are: I’m a little girl or little boy. Then they use that information to look out into society and say: How does society treat other people who are like me? If I’m a girl, I pay special attention to how society treats little girls in this society and I discover that they treat girls as if they don’t like math or don’t do robotics or computer programming or engineering.
And pretty soon, when you pick up that information, that stereotype, you begin to apply it to yourself and it can limit your aspirations, what you think about your future and so forth and so it has deep implications.
Have you found out where this information is coming from? Television? Peers?
Peers are a big influence, parents are a big influence, even teachers can be. So that’s part of what we’re not trying to study in the laboratory: What are the sources of the stereotype? We started with [the question]: What age do the kids catch them? And we were quite surprised that these were stereotypes about math as early as second grade, before they learned their multiplication tables. So this shows that the stereotypes are impacting the kids quite early.
We also did some studies about kids catching social bias: Who is a good person or someone you want to associate with? And who you might think is a bad person? Kids catch bias from watching adults too.
The study we recently published about catching social bias was really interesting. We had two videos in front of four- and five-year-old children, and the person on the left, let’s say, would wear a dark green t-shirt, and the person on the right would wear a different color t-shirt and then there was an actor in the middle. And the actor would turn to the person on the left who had the green t-shirt and say, “Oh, hi! How are you?!” And hand them a toy and act enthusiastic. And then the actor would turn to the right with the person in a different colored t-shirt and say: “Oh. Hi. How are you?” [in a distrustful voice] Then, they would hand them the object, but they’d do it in a very hesitant way. So the language was the same to both targets, and the action was the same, but they were done in a different manner. One was done very upbeat and pleasant and the other one was done as if you were hesitant to interact with them. The little four and five year olds simply watched this film—we didn’t give them any instruction at all—they simply watched. And then we gave them a toy, and said, “Who would you like to hand this toy to?”
And the children said, overwhelmingly, that they wanted to hand it to the person that the adult had acted favorably towards rather than the one the adult had acted unfavorably towards.
When we had them do little exercises about who they would learn from, the children said they would learn better from the person that had been responded to in a pleasant way rather than an unpleasant way. So it was if the children were picking up our biases simply by watching our behavior.
Now, that’s an interesting thing because when we’re in the supermarket lines with our children and someone comes up behind us in line and we grab our child’s hands and clutch them toward us or act like you want to avoid that person, the child is picking up that cue, looking at who that person is and realizing: Mom does not want me to associate with that person, that maybe that’s a bad person. If you speak in systematically unpleasant tones of voice to one type of person rather than another, your child is watching that. Your child is observing biases from observing you. So it’s not only what you say to your children: "all people are created equal" or "you should be nice to everybody." Your children are watching how you treat others and taking information about that. That’s why this little study was so important.
As a matter of fact, we did a follow up to that study where we introduced the child to others who wore the same t-shirt colors of the first adults they saw. If one person had a green t-shirt on and the child had seen the green t-shirt person responded to positively, when they saw a friend of that person who wore a green t-shirt, or a friend of a different person who wore another color t-shirt who was responded to negatively, the children interacted with these new people in the same way. So they would treat people with the green t-shirt as positive people that they can interact with and people who wore the other color t-shirt as a bad group of people who they shouldn’t interact with. So children generalize. It’s not only that you reacted to this particular person in a negative way, but children very quickly realized to act in a negative way to that whole social group, that whole class of people.
Are you able to go from this example about people in green t-shirts to people of other races or classes etc.?
That is the exactly the interesting point. So we are interested in this laboratory demonstration that children pick up social biases from watching adults and they generalize to a whole social class.
We’re obviously thinking of this as an unconscious learning mechanism that children are using every day to learn about how we socially group the world and we attach attitudes to that world, like: These are good people, these are bad people. And the children are picking up that information from watching us.
So that makes this line of research very important and significant in today’s world because it's about what children are picking up from adults about prejudice and how we respond to social groups. We did not realize that children as young as this were carefully paying attention to how we treated others as social groups and would assimilate that information to themselves.