131: Implicit Bias with Dr. Mahzarin Banaji
Explicitly, nobody really believes in gender stereotypes anymore, but when we look at the world, and who’s where and how much money people make, and so on, it still seems to be there. And the answer to that is yeah, because it’s there. It’s just not something we say. It’s more of something we do.
-Dr. Mahzarin Banaji
What is implicit bias? Do I have it (and do you?)? Does my (and your?) child have it? And if we do have implicit bias, what, if anything, can we do about it?
Join me in a conversation with Dr. Mahzarin Banaji, former Dean of the Department of Psychology at Harvard University and co-creator of the Implicit Association Test, for an overview of implicit bias and how we can know if we (and our children) have it.
This episode will be followed by a second part in this mini-series where we dig deeply into the research, where results are complex and often contradictory. Stay tuned!
Jump to highlights:
- (01:00) An intro of Dr. Mahzarin Banaji
- (02:58) What is implicit bias?
- (07:48) Differentiating bias that you are aware of and bias that you aren’t aware of
- (08:56) Describing the Implicit Association test
- (18:11) What the research says about where implicit bias comes from
- (24:50) Development of group preference from implicit association
- (32:18) Group bias and its implications towards individual psychological health
- (40:44) What can be done to potentially prevent implicit biases from developing?
- (46:56) Some good progress with society’s bias in general and areas that need working on
Resources:
Click here to read the full transcript
Jen 00:02
Hi, I’m Jen and I host the Your Parenting Mojo podcast.
Jen 00:06
We all want our children to lead fulfilling lives but it can be so hard to keep up with the latest scientific research on child development and figure out whether and how to incorporate it into our own approach to parenting. Here at Your Parenting Mojo, I do the work for you by critically examining strategies and tools related to parenting and child development that are grounded in scientific research and principles of respectful parenting.
Jen 00:29
If you’d like to be notified when new episodes are released, and get a FREE Guide called 13 Reasons Why Your Child Won’t Listen To You and What To Do About Each One, just head over to YourParentingMojo.com/SUBSCRIBE.
Jen 00:42
You can also continue the conversation about the show with other listeners in the Your Parenting Mojo Facebook group. I do hope you’ll join us.
Jen 01:00
Hello, and welcome to the Your Parenting Mojo podcast. Today we’re going to look at the topic of implicit bias. Now I’ve been thinking for a while about running a series of episodes on the connection between our brains and our bodies because I’ve been learning about that and the wisdom that our bodies can hold and wondering, well how can we learn how to pay more attention to our bodies? And then I started thinking about intuition. And I wondered, well, how can we know if we can trust our intuition? What if our intuition is biased? So I started looking at the concept of implicit bias and it became immediately clear who I should ask to interview Dr. Mahzarin Banaji. Dr. Banaji studies thinking and feeling as they unfold in a social context with a focus on mental systems that operate in implicit or unconscious mode. Since 2002, she has been Richard Clarke Cabot professor of social ethics in the Department of Psychology at Harvard University, where she was also the Chair of the Department of Psychology for four years while holding two other concurrent appointments. She has been elected fellow of a whole host of extremely impressive societies and was named William James Fellow for a lifetime of significant intellectual contributions to the basic science of psychology by the Association of Psychological Science, an organization of which she also served as president. Along with her colleague, Dr. Anthony Greenwald. She’s conducted decades of research on implicit bias and co-authored the book Blindspot: Hidden Biases of Good People.
Jen 02:21
I should also say that there are a lot of issues that we only got a chance to skim over at a fairly high level in this conversation, which I’m recording this introduction afterwards, because Dr. Banaji was quite pressed for time. And I’m planning to release an episode that follows up into these issues and dives into them at a much deeper level soon. So please consider this part one of a two-part conversation with you.
Jen 02:42
Alright, let’s go ahead and get started with the interview.
Jen 02:45
Welcome Dr. Banaji. Thanks so much for being here.
Dr. Mahzarin Banaji 02:48
Hi there.
Jen 02:49
So I wonder if we can start out by understanding a bit more about what implicit bias is. I hear it all over the place, and can you help us to just define what that is, please?
Dr. Mahzarin Banaji 02:58
Sure. So implicit bias, quite simply, is a tendency in every human being to favor one individual over another, one social group over another, and to do so without conscious awareness, or without the ability to be able to exert conscious control over the judgement that one is making. So let’s just break the phrase down into the two words that constitute it. The word implicit and the word bias, okay? Bias, what is it? It’s simply for us a deviation from neutrality, it is privileging one option over another, right? If I say I prefer blue to red, I’m biased in favor of blue, and not in favor of red relatively speaking. So that for us is what a bias is. And implicit just means that that favoring of red over blue or blue over red is something that I’m not even aware of. It’s just that when I go into a store, I pick up clothing that is all blue rather than red. That put together tells us that implicit bias is a deviation from neutrality in ways we ourselves would not be happy to see ourselves doing. If it’s in the domain of color, who cares whether I prefer blue to red or red to blue but imagine now that it’s not blue and red. Imagine that it is a native born child in the classroom, and an immigrant child in the classroom. And even though I as a teacher believe very much that both should be treated equally that is to say, if they do something good, I should reward both equally. If they did something that’s bad behavior, I should reprimand them equally, I should encourage both equally to pursue new things in their lives. I should support both of them equally to meet their goals and so on. A deviation from neutrality would mean that I’m doing these things both the good and bad things in order to teach a child something, that I’m doing them selectively, that I’m doing more for one category over another. So that’s all biases. And teachers are so well intentioned, just like parents, what they want is the best for all of the kids in their class. And so when we discovered that a teacher may not be aware but is systematically calling on certain people in the classroom, or saying, “Aha!” or “Good idea!” to some rather than others, then we would say it’s implicit. And as you can imagine, teachers, by and large, are very good people and so when they’re biased it is almost always without their awareness.
Jen 05:41
Okay, so is this lack of awareness perspective, that’s really the key, then?
Dr. Mahzarin Banaji 05:45
That’s exactly right. And the reason this is interesting is because if you look in almost any society, but let’s just take American society or the Western world or whatever, some large group of people, you will notice that explicit forms of bias have been coming down, at least in what people say on a survey. If you say to a teacher, “Do you think that native born children are just inherently smarter than immigrant children?” The teacher will likely say, “No, I don’t believe that that is the case at all. I think all children are talented in all these different ways.” So if you measure it explicitly, if you say, “Tell me is this immigrant kid better or worse than this native one kid?” You will not see any evidence of bias. But when you sit in the back of the classroom, and you just measure what the teacher is doing, who the teacher looks at who the teacher says nice things to who the teacher calls on, and you see that there is a systematic difference, then we say we must become interested in this, because the child is experiencing these good and bad things the teacher is doing, but the teacher has no awareness. And think about the child. What does the child think is going on? The child might think I’m bad, or I’m good, right? And that’s why we should be interested in both kinds of measures, what people say to us directly, and also what they may not be able to say because they don’t think they’re that way.
Jen 07:10
Okay, I’m wondering if we can just pull that apart a little bit. And we’re sort of using teachers as an example. But this could just as well apply to managers or anyone in any situation. I can understand if a researcher were to come and say, “Do you think that native born children have different capabilities than immigrant children?” Then, I understand the correct thing for me to say in that moment is “No, of course not.” But I may still be thinking that and I may have awareness that I’m thinking that. So I’m trying to understand the difference between an explicit bias that I know it’s not a socially correct thing to say, and implicit bias that I might not be aware of, how can I parse that difference?
Dr. Mahzarin Banaji 07:48
Yeah, it’s a great question. For now, I would say, an explicit bias is something that you know, even if you don’t say it to other people, but you know. So let’s say that I believe that boys are better at math than girls are. But I’m not going to say that because I don’t want my girl students to hear that or feel bad about that. And so I’m not going to say it, but I think it and I’m able to consciously put those words together in my mind, boys are better at math than girls, I know that. As long as that is the case, we would say it’s consciously accessible to you. Your mind is capable of saying A is better than B or whatever and to do that to yourself, at least. What makes it implicit is when you say boys and girls are equally good at math, I believe it. You never say to yourself; boys are better at math than girls. But if we look at some other aspects of your behavior, who you spend more time teaching a difficult math problem to, etc., then we would say it is.
Jen 08:56
Okay, perfect. Thank you for helping us to understand that distinction. And so then I wonder if we can go from there into the Implicit Association test, which is something you’ve spent a bit of time working on over the years. Can you tell us what that is? And how does that measure implicit bias?
Dr. Mahzarin Banaji 09:11
Yeah. So as you can imagine, your listeners and you will easily see what the problem is. A person, if asked genuinely and truly says, “I have no bias.” And when you measure their judgments and actions in some way, you’re seeing a systematic effect. How do you measure that? When the person themselves is saying no? In psychology, we’ve relied for over 100 years on what we call explicit measures. If you want to know something about what a person is thinking, ask them. How stressed are you feeling? Okay. Well, sometimes maybe I can tell you that I’m feeling stressed. But there are lots of studies where people like me say that they’re not stressed at all and you’ll see me breaking out into hives or something, which is a response to the stress. Now you have to have a measure of those hives, you have to be able to measure skin complexion, you know, when one is stressed and not stressed and say whether the person knows it or not, there is some physical response that we can measure. What can we do when the response we are looking for is locked up between our ears in a box that is not easy to penetrate? So the first thing to do is to just imagine the difficulty of trying to even track anything implicit and the measure that you mentioned, the one that we are most familiar with, and the one that I believe today is the dominant measure of implicit cognition in the science as a whole is the Implicit Association test, and I’m one of three co developers of that test. The test has a very simple assumption that underlies it. The assumption is that when two things in our experience have come to go together repeatedly, they’re joined in time and in space, let’s say, that they become one for us. If when I see bread, there is usually a bowl of butter, bread and butter come to be associated when I think bread, butter comes to mind more quickly than some random word like water, or couch or something like that. So that’s very easy to understand and neuroscientists, actually, in order to teach people how neurons in our brain fire after having learned something, they teach it to us by the phrase, if it fires together, it wires together. And we use that when we teach how learning occurs. By firing together, it means in the same moment, if neurons for bread and butter becoming activated, your brain learns that there’s something about these that go together, and that’s how we learn everything. We learn, you know, that mother and father are a unit, like bread and butter, but we also know certain experiences we have in the presence of something. When I see flowers, I feel happy, and flowers becomes associated flowers almost become synonymous with good, even though the words that we might use to capture goodness have nothing to do with flowers. So you know, not just words like beautiful or peace or joy, but if we use words like you know, angel, or satisfied, or whatever that have no semantic relationship to flowers, those words should become more easily accessible in the presence of flowers because our experience has made them repeatedly be associated. Unlike insects, when we think of an insect, we think yucky, I mean, unless you’re a five-year-old boy, yes, I will put them aside for a moment. And say that, yes, that may happen. But even young boys, when they take our test, if it’s flower or insect, they have learned that in our culture, flowers are good, and insects are bad. So all we’ve done is made a test that measures the strength of association between insects, and bad and good flowers and bad and good. Okay, that’s what the test is. And now we can begin to go beyond the test by keeping the test logic identical. If I have you look at flowers on a computer screen and press a key – left key when you see flowers on the right key when you see insects – very easy for you to do that. Now I’m going to say okay, not just flowers and insects, but words are going to pop up on the screen words like love and peace and joy, good words, or bad words like devil and bomb and war and things like that. And your job is to use the same key to say flower or good words, and the same key, now a different but the same key when you identify an insect, as having appeared on the screen, or a bad word is having appeared on the screen. Now, if flower and good have truly become one in our minds, this should be a very easy task. Left key for flowers left key for good. Right key for insect right key for bad, right? That’s easy. But now let’s switch. This is the moment in the test when people groan because I’m saying to them left key for flower and left key for bad things, right key for insect and right key for good things. And they can’t do it. By can’t do it I mean, they can, but it takes them a whole lot longer to do this and they make many more mistakes when they do this. I show this bias, you show this bias. You know, even entomologists who study insects and love them, show it but to a lesser extent than we do. So they acquired the cultural thumbprint that says insects are not as good as flowers, but because they love insects and they work with them. They show a lower anti insect bias.
Jen 14:51
Fascinating.
Dr. Mahzarin Banaji 14:52
Okay, so now, the logic of the test should be very clear to people, and they will mostly agree in fact, everybody will agree that this is a decent measure of whether we like flowers or insects. And when the data come back and tell you, you have a strong preference for flowers over insects, people nod their head and say, “Yes, I do.” There’s no quarrel with the test. The quarrel with the test emerges when we replace insect and flower with Black and White faces, Asian and White faces, fat and thin people, people who are Native Americans and European Americans. And sometimes when we even change the good and bad words, to be things like bad and good things like weapons and musical instruments or something like that. Is it easier for me to associate, you know, certain groups with bad things and certain groups with good things? And the startling result is that for people who have, and I would say, genuinely have no explicit bias, now, I can’t say what your explicit bias is, because you may think that you think that you know, male is better than female, but you may not be willing to tell me, but I’ll take myself, I know that explicitly, I do not believe that Black is bad, and White is good. I know that for sure because it’s me, and I can tell myself the truth about what I consciously think. And yet for people like me, who seem to have no explicit bias, this test throws us a curveball, because it demonstrates that people like me are not able to associate good with Black as easily as we can associated with White. And when that happens, it’s troubling to us. It’s troubling because it doesn’t feel like the test is telling us anything true about ourselves. I don’t blame people who say what a stupid test is just telling me, you know, it’s just all complete lies. I felt the same way when I first took the test. I thought, “What’s wrong with this test?” Obviously, I’m the great Mahzarin I’m not biased, so if the test is telling me I am, something’s wrong with the test. And you know, a few minutes later, I came to my senses and I realized it’s not the test that’s the problem. It’s my head that’s the problem. I have accumulated all this learning two things have gone together, it fired together, it wired together.
Jen 17:22
Okay. All right. Well, thank you for that. And I think there’s a lot of different pieces to come up from that, that we’re going to get to in due time as we go through the conversation, but I’m wondering if maybe we can start with “Where does this come from?” Because in preparation for this interview, you sent me an epic paper that you had written with one of your grad students, and the reference list alone was enough to make me weep when I saw it, and realized I was probably I need to read most of those.
Dr. Mahzarin Banaji 17:50
Which paper was that?
Jen 17:51
The one with Dr. Charlesworth on, I hope I’m remembering her name, right?
Dr. Mahzarin Banaji 17:56
Yeah. Charlesworth is writing, but she’s written too many papers.
Jen 18:00
You’ve written a lot with her. But the idea of the assessment of where implicit beliefs and attitudes about gender and race and language come from. So…
Dr. Mahzarin Banaji 18:08
Oh the language paper.
Jen 18:09
Yeah.
Dr. Mahzarin Banaji 18:09
I got it.
Jen 18:11
Yeah, sort of the overview of what does the research say about where all of this comes from in all of these different domains of gender, race, language, and age, and so on. So wonder if you can walk us through what we know about where these implicit biases come from?
Dr. Mahzarin Banaji 18:26
Yeah, so whenever we talk about where does something come from, in this domain, you must always think about it in two buckets. The data that we know, female is nice, male is strong, whatever those beliefs might be. They’re obviously coming from somewhere, right? And that’s usually outside of the human in some sense first. When you’re born into the world, you don’t know that you’d know in a few days that mom is nicer than dad because you’ve been attached to mom. And infants, by the way, do notice very, very quickly, that is they know, they prefer other females if their primary caregiver is female, and male, they prefer other males if their primary caregiver’s male. So they’re clearly learning that. But the very fact that they’re learning that female is good or male is good, depending on their experience tells us that those experiences are coming to them from the outside. But there is a machine inside. That’s the second bucket. It’s not like it takes a long while for these things to get learned. There is a learning machine in our brain is… it’s like a sponge for anything new that comes to be paired with each other. We’re seeking those associations and when we see them, we form immediate hypotheses. “Oh, a person who looks like this is a nice person.” They feed me, they take care of me, they feel cuddly and warm, they stroke my cheek and some other group that has never done anything bad to you, but just hasn’t done anything good is not associated with that, right? So, when we think about where this comes from the specific content, female is soft and nice feeds me, etc. Male is strong, you know, can carry large objects, whatever those are, those are obviously being learned, it is not something that the brain knows on day one. But the thing to focus on is that it’s learning it extremely fast. So fast that as I said, within the first, you know, two months of life, babies are showing preferences for dark skinned people if their caregivers are dark skinned, and for light skinned people, if their caregivers are light skinned.
Jen 20:45
Okay, and so there’s sort of a shift that’s happening in there. And I want to make sure that we see it happening. In the very early days, there’s a noticing which researchers are measuring things like looking studies, like how long does a child look at one person versus another person, one picture versus another picture.
Dr. Mahzarin Banaji 21:02
Yeah, I don’t think looking at any, as really anything, like just looking. Looking almost immediately is going to also be reflecting either a preference, or that it’s something startling and unique and I better look at it, because this doesn’t sound like the usual thing. So looking time is telling us more than just here are two things and I’m going to look at one over the other.
Jen 21:24
Okay. And that’s sort of linked to the categorization that’s happening, right that one particular person has these features and here’s another person that has these features. And this person brings me milk and stuff that I like, and this person looks sort of like this person. And so we start to lump these together. Is that right? That that’s how the categories form?
Dr. Mahzarin Banaji 21:46
Yeah, when I say that, babies very early starts to show a preference for others who share the gender of their caretaker, that’s exactly what’s happening. If one woman is good to you, your brain apparently thinks that’s more than enough data. And you will generalize from that to other women. And if you have a mother and grandmother, it’s over now all women are going to be good, right? Because two of them have already shown that they did not harm you. So yes, I think the question of categorization and then forming liking for one, and the other is too complicated to get into, because there are many different views about all of this but the important thing is, yes, very early babies can put things into groupings. And very quickly, and this is the interesting part, the fact that they can see that things are different is not so surprising, but what is surprising is how quickly beliefs and affects come to be associated with that. How quickly things become good and bad. That’s the part that I would be interested in.
Jen 22:48
Yeah. And so why do we do that? Like, why does the good and bad thing happen? Where does the preferences part come from?
Dr. Mahzarin Banaji 22:54
Because it’s so adaptive. For babies that didn’t do this, they would just die. I mean, imagine that a baby fails to distinguish between people who take care of them and people who harm them. A baby like that would not survive a day because they would approach people who are likely to harm them. So for a baby, to not go in the direction of people who look unfamiliar, is a very protective mechanism. For the baby to learn very quickly that when somebody has not harmed me, others, like that person are not likely to harm me is a decent hypothesis. It will pan out in their little universe. So that’s where it comes from. It’s not for nothing, that they’re, they’re not stupid, right they’re, this is how humans evolved. Humans who did not have the capability to distinguish between categories and assign a positive or negative value to those would have died out in our evolutionary history. They didn’t survive long enough to have their own children, so that they would be amongst us people who don’t make those distinctions. This was so important that it is now a basic human quality because only those…
Jen 24:03
And it happens across categories, like gender, and race and language and age. Those were some of the ones that you covered in your paper,
Dr. Mahzarin Banaji 24:11
And tables and chairs and cups and saucers, and apples on everything.
Jen 24:16
Okay. And then I was curious to see that boys undergo a shift, girls undergo less of a shift as they get a little bit older and go from and start to see female with good and male with bad on a version of the Implicit Association test that’s designed for children. And so instead of preferring their own group, because I think the tendency is to prefer own group and to for boys to say, “Oh, boys are better than girls.” And girls say, “Well, girls are better than boys.” And then over time, it becomes sort of a female preference. What’s behind that?
Dr. Mahzarin Banaji 24:50
Yeah, you’re asking a really important question. So let me just back up a little bit and say, you’re exactly right. In group preference, we see in some ways is ubiquitous. You can go anywhere in the world and you won’t be surprised when you see that people like their own group better than they like other groups. If it’s something like the Red Sox and the Yankees, it’s very explicit, I can actually scream all kinds of epithets about the Yankees considered socially completely. Okay,
Jen 25:22
I enjoyed the research paper on that particular…
Dr. Mahzarin Banaji 25:25
I can wear a T-shirt that says, “The Yankees suck.” And even though they don’t, and they’re far better than the Red Sox, it is okay for me to do that. So in group preference, starting with something as clear and as socially acceptable as commitment to a sports team, and hating the rival team, starting with that, and going down to other kinds of groups. American, is American better than European for Americans, Americans are better and for Europeans, Europeans are better. So let’s just start with the assumption that in the psychological research literature, this has been shown thousands and thousands of times in every possible way, that in group preference is strong and ubiquitous. When you find that not being the case, you have to get interested as a scientist, how can it be that something as pervasive is failing to appear? And we see it more on the IAT on the implicit tests. So on the explicit test, if you ask Black people, which group do you like more White or Black, they will say with lots of passion, we think Black is better than White. When you ask White people, which group is better White or Black, they will say their own group is better, but a little more modestly, because they know their history. They know that it’s not so cool for them to say White is better than Black. So they will, they will still say it but in a much more modulated way than African Americans will. So that’s what they say. Now let’s go to the IAT and we’ll use both examples race and gender here. On the IAT, White people show strong preference for White over Black, and young children, six-year-olds, and even younger, we now know, show that same preference. So they’ve already learned by the time they’re four, we know it can be seen on lots of complicated tests that they prefer, just like the adults of their group. So why do adults and White kids look the same to each other, they reflect each other and they show clear in group preference much stronger in group preference than they might say, on a scale if you ask them the question, which group do you prefer, and so on. Now, let’s go to Black Americans. If the world were fair and equal and nonhierarchical, Black Americans should if 70%, or 75% of White Americans show White preference, then 75% of Black Americans should show Black preference. And if that’s steam, then we would say this is symmetric, opposite but symmetric, right. But that’s not the case. 40% of Black Americans show in group preference 40% show out group preference that is to say they prefer White to Black and 20% are neutral. When you collapse the data for Black Americans, you see a graph where you see roughly half the people on the side have White preference, and the other half on the side of Black preference. And young Black children, six years of age will show the same result. They have learned. This is to me one of the morally difficult results that by that age, a Black child knows enough that even though they might say I like Black kids, they do not show that preference on the IAT. So what is the IAT and the value of implicit measures is that they are capturing the thumbprint of the culture on our brain, not what we think we ought to say. And now you start to see the same thing with gender, but in a way that will perplex people. In the world there was this wrong belief that because women are oppressed, because women have been subjugated by men, because women have had many fewer rights than men, that women will be seen as bad. Well, women are not respected or seen as strong or competent but they are loved because a very defining meaning of female is mother in every culture. The basic meaning of female is somebody who is nice, who will cook for you, will hug you, is warm, smells good. All of those things are associated with mother. So if you give people an IAT – male, female, good, bad – girls will show in group preference just like it’s supposed to be. Girls would like girls, but young boys who have learned that mother is wonderful and nice they will show a lower preference for female but they won’t show just like Black people don’t show a strong preference for Black over White, boys won’t show as strong a preference for male or female, because they’re members of that group. So they like the group male, but they too have learned that in our culture of the two groups, male and female, the good one is mother. Not necessarily the competent one, the strong one, the one who brings in the money, the one who, you know is going to be able to kill off the enemy for us all of that, not that. So what you now see is that it is silly for us to assume that any particular group is all good or all bad, it depends in which category you’re looking. There’s a fundamental warmth that we might feel towards some people over others. And that’s one dimension on which we like or dislike certain groups. But as Susan Fiske, a psychologist has shown there is a second variable that does not run parallel, it runs almost in opposition to the warmth one that are loved but seem to be not so strong. Or you could be seen to be strong, like father, like CEO, like feminist women, these groups are seen as strong, but not necessarily nice. So that’s what you see and the amazing thing is that what we see in adults we’re seeing in young children, and that’s the important message here that a lot of parents struggle with this. I have had numerous parents write me, email, tell me, please don’t tell anybody this but you know, my child said something, something terrible, racially. And I have no idea how they could have gotten that because it’s never said in our home. It’s never said in the school that they go to. How did this happen? And I say to them, that it’s because they have a very impressive view of their own influence on their child. Their child is influenced by what it sees in the world. And if they go to a school where there are Black and White kids and the Black kids, you know, have a lunchbox that is more beaten up than the White kids’ lunchbox, that might be enough to say that’s not as nice as this other nice, shiny thing. And that’s all it would take to figure all this out.
Jen 32:18
Okay. And so two things to pick up on that. Firstly, the implications for Black children and even Black adults have this sort of two humped curve as it were, with half of the people preferring the outgroup. And with all of the messages that our culture sends about the outgroup’s superiority in this case, what implications does that have for the psychological health of the people who are in this group?
Dr. Mahzarin Banaji 32:46
So I’ll tell you two things and they will I hope both be interesting. The first thing is that as soon as you combine what I just told you about in-group and out-group preference, that in-group preference is robust and strong, in dominant group members, but it’s not present in members that are minority groups or disadvantaged groups. We begin with that result and then we can actually make the effects even worse by saying they’re not even the same size statistically, in every culture. If White Americans are about 60% of the population, but Black Americans are only 10%. Just count up, just do simple math. How many good things is a White person likely to get being thrown towards them in a given day? Oh, 60% of their population, or other people like them who show preference for them in their group, African Americans, even if they showed strong preference for their own kind, would only have that be available in the 10% in that group. So statistically, it makes the result worse, even worse. So I think that this is a battle that has to be fought inside of each person. What am I to do? I love myself; I want to love myself, I want to love my home and my people and my thing, but I know that they aren’t good. The results that I will give you next is a very shocking result. When you measure self-esteem – how much do you love yourself? – and even when you use measures that are implicit, associate good things to yourself, we find that African Americans have the highest self-esteem in our culture, followed by White and the lowest is in Asians. So how does this happen? How does a group that is discriminated have the highest level of love for self in some ways? And I think it’s because Black kids all the way to Black adulthood are constantly being tested. Self-love is not something that can be set aside. We love ourselves. We have to protect ourselves. That is so powerful that Black Americans have to work harder. This is my conjecture, that the reason they show high self-esteem is because they don’t have the luxury of deriving it from their social group. White Americans can feel good because other White people’s achievements are theirs. Black Americans have much less of that to lean on. And so it has to be done based on yourself, I have to be the good one, I’m going to have to do the good thing. And I think that it is this kind of constant mental work that goes on to make sure that the self is protected, the self is seen as good, that leads to an almost better developed self-love that in African Americans is coming from the hard work they have to do to make sure that the goodness they associated with themselves is coming from their own performance from their own being good, or what that might mean because it’s not going to come from just, you know, having it showered on you by others in your group, or by even having your own group members show it to you. But this is a very complicated story. But the result is not the result that Black Americans, so this is a contradictory result, because we used to think that, you know, if your group is discriminated, you will have even had a word for it, we call it you know, self-hating x, we would say, well, that’s it’s not so simple, psychologically, we should just be aware that love of self is so primary, that we will find a way to love ourselves. And at the even at the expense, even when we know that the group we belong to is not so good, culturally speaking.
Jen 36:37
Okay, then that should sort of an additional layer of the next thing that I wanted to follow up on, which was when you said that parents come to you and say, you know, my child said something inappropriate and where did this come from? And you said, Well, this is sort of the culture exerting its thumbprint on your child’s mind. Of course, there are also many studies showing that children at very young ages, if a researcher asks them and presents them with a picture of a Black baby and a White baby, a doll, or an actual child picture than both of those Black and White children, who are the study, participants will say, “h, yeah, the White baby is nicer. The White baby is cleaner. The White baby is better in some way. And so how can we know that these results are from implicit biases, rather than explicit biases that the child hasn’t yet learned that society says that they need to cover up because it’s not socially acceptable to say this?
Dr. Mahzarin Banaji 37:29
Yeah. So in our very first study that we did, using an implicit measure, it was done with Andy Baron, who’s a professor now at University of British Columbia, we were the first to just try and IAT with young children. And so we first asked six-year-olds, ten-year-olds, and adults to just tell us, you know, here two people, which one do you like. Six-year-olds are the most honest, White six-year-olds, 90% of them tell us I like the White one better than the Black one. By ten, they have started to learn that that may not be so cool to say. So they still show fairly decent in group preference when they are asked, but explicitly, they’re not as likely to say the same. And when you ask adults, you see, nobody said Oh, because they say half the time White is good. And half the time Black is good. And that’s a good measure of how civilized these adults are, at one level, right? At least explicitly, I’m going to take them at face value. They’re saying my values are such that I prefer both equally. When you look at the IAT data, of course, there are no such differences. six-year-olds, ten year olds and adults all show strong in group preference. All right. So kids are in fact, honest, they will tell us what they think and in fact, young Black children are being honest, when they say the White doll is better, because they know that in their culture, that’s the case. There is one videotape that’s been made of the Clark and Clark doll study, which is the one you bring up originally done in the 1930s by Kenneth and Mamie Clark, a study that was cited in the important legal decision Brown v Board of Education, and that’s the original doll studies. But somebody did a version of it not in a systematic way as the Spencer’s have done, but just sort of videotaping Black kids, as they’re shown Black doll and they picked the White doll when they’re asked which one is good. The kicker is when the experimenter then says, which one is more like you. And you can see a child’s face almost turn adult-like in its demonstration of conflict when it very slowly and hesitantly points to the Black doll to say that’s more like me. What I’m saying is that, yes, they know that and at that age you know, who knows what it’s doing to self-esteem? What we do know is that exactly that process must be working its way, all over the place as they grow older, so that by the time they’re reasonably old, they’ve learned that I am good. My group is not good. And they figured out a way to derive self-esteem. But this is not studied very well. So I don’t want to go too far into it but I can tell you that’s been well studied is that Black people show a fairly high self-esteem that we know. And it’s a puzzle as to how is it possible given our simplistic belief that people who belong to disadvantaged groups will dislike themselves and so on? It’s not so simple. We can’t say that.
Jen 40:44
Yeah. And so then, then we could, of course, come to the issue of well, what, if anything, can we do to potentially prevent implicit biases from developing, and I’m just thinking to the study that you just published this year, again, with Dr. Charlesworth on the massive study of gender stereotypes and natural language. I mean, this is enormous in scope, it’s millions of words of assessment that would never have been possible to do before computers that could analyze such high volumes of data. And so when we think about the findings from that study, which are the words associated with gender are everywhere, in natural language interactions in movies, and everything that we interact with our children, what can we do when all of this stuff is so baked into our culture to potentially prevent implicit biases from developing?
Dr. Mahzarin Banaji 41:40
Yeah, so let’s start with the answer to the parents who say, I never say to my child, anything that’s gender stereotypic. I’ve known parents who will doctor their children’s books to put breasts on truck drivers so that their children don’t learn a standard. And still, the kid somehow seems to know the right answer, truck drivers don’t have breasts. Let’s start with that. On the one hand, the parent says that, and then let’s look at just parent-child conversations. So in the work with Tessa Charlesworth one data set that we used to look at this using this technique called word embeddings, which words go with which other words which words are semantically more connected, if I say to you, kitchen mitt and mother are going to be more associated than kitchen mitt and father, you would have no trouble understanding that that might happen. If I say to you, baseball mitt and father are more associated than baseball mitt and mother, you would not be surprised. That’s what the data would show that in our language, when kitchen mitt appears, it’s usually what she mother etc., when baseball mitt appears, it’s more with pronouns and names that are male. What parents are telling us is that we don’t explicitly say girls are, you know, not competent, and boys are competent, but parents need not say that. Right? If in our culture, a person who’s working on a car in their garage, is somehow culturally seen as more valuable than stirring soup on a pot in a kitchen, and mother and kitchen are associated and father and garage and cars are, then automatically children are learning that fathers are better than mothers, because what fathers do the work they do is better than the work mothers do. Mothers are nicer than fathers, but mother’s work is not so good. So when you look for stereotypes, of associations to in our culture, things lower on the hierarchy, you know, helping and following rather than building and leading, you see that in the conversations parents and children are having, so this is the brilliance of the data set test out worked with, she has one data set of just parents and children talking. In the course of normal life. We know the age of the child, the gender of the child. We know a few things, but not much else, because we only have audio recordings. So we subject these audio recordings to this analysis and we find that the evidence for use of gender stereotypes in both parents’ language and the child’s language is just as robust as it is in TV shows and in music or in encyclopedias or in books of fiction or nonfiction books or whatever. So that’s the important data. Stereotypes in the language when we’re speaking, are robust. They’re seen equally across many different language corpora, including parent-child conversations. And I think for your parents, this should be an interesting idea that they’re right when they say I never say that girls can’t work on a car. No, they never said that. But in their life, they’re demonstrating that by having the male in the family do that and the female not do it. So they don’t ever have to say it. It’s seen in the behavior of these individuals. And that’s how children pick it up because it would be silly not to be a good learning machine that’s trying to learn what goes with what.
Jen 45:16
And then I mean, it seems to have so many links to patriarchy as well and the idea that they use you said helping and following are not good. And creating and leading are good. And these are stereotypically male and female traits.
Dr. Mahzarin Banaji 45:30
So older feminists often argued that we’ve got to get women into these high positions, so that they will be associated with all those values of, you know, respected work, and so on. And I think a newer breed of feminists might say, no, we have to make sure that housework becomes seen as important. That, you know, in cultures where it is paid work, or whatever, that’s a way of a culture signaling that it really believes that the work mothers do is just as important, in fact, more important because there are the schools, I’m told in Denmark, where to just send your child to school, you would be paid, you know, 700 something a month. And that’s saying, you know, we care enough about protecting mothers time and sending kids properly to school, that we’re going to pay people to do that. Until that happens, I don’t think, and that’s why people are perplexed, why isn’t that explicitly, nobody really believes in gender stereotypes anymore. But when we look at the world, and who’s where and how much money people make, and so on, it still seems to be there. And the answer today is yeah, because it’s there. It’s not, it’s just not something we say, it more of something we do.
Jen 46:40
So then it seems to me that the trick here there’s a trick is to firstly, have the conversations with your child about the value of these different professions and roles and careers. And then secondly, is taking on non-traditional roles and responsibilities to the extent that you want to do that.
Dr. Mahzarin Banaji 46:56
Yeah, so Tessa’s tests, the ones we analyzed are twofold: we take a gender test, where we measure how strongly do we associate, male and female with the categories of career or work-related things, and then home-family related things. It’s a robust relationship. You know, many, many people, the vast majority showing association between male and career female and home, including women at the same level as men do, so men and women do not differ in this bias. Women just as much as men, if not a little bit more associate female with home and male with career. But in the 10-year period that we’ve looked at roughly 2007 to 2016, we see that this bias is coming down as a function of time, but also a function of age of the child. And that’s because that is the world, right? I’m an old person. So I learned what those are. But my world today is different than my world in the late 50s. And so I am changing. But a kid who was born only 10 years ago, begins with a world that is already very different than 1950s and they’re changing. So what’s great about Tessa’s data is that she doesn’t just look at absolute levels of bias. She looks at change over time. And even in attitudes towards gay people. Everybody’s changing as it turns out. We’re all becoming less and less anti-gay. Conservatives and liberals, men and women, older people and young people, the coasts, the middle, rich, poor, educated, less educated, everybody is becoming less anti-gay. But two groups are becoming anti-gay faster than everybody else. And that’s young people and people who identify as liberal, those two are almost already close to neutrality. When we talk about bias, we said, a bias is a deviation from neutrality. If you like Black and White equally, or if you’re like gay and straight equally, your IAT score would be zero because whatever bias you’re showing here, you’re showing there, and when you subtract one from the other, what’s left is zero, which means there’s equality in whatever you feel it’s equal. And what we’re seeing is that zero point is being reached for today already by young people and by liberals. They’re not showing in anti-gay bias. So those data make us very hopeful. We just have to figure out a way of bottling whatever is going on with sexuality, and then see if it can be applied to changes in other ways that are not changing as fast. So race is changing anywhere as fast. And age bias. Your older parents or grandparents might be interested, although it’s a sad result, that one of the most robust biases that we’ve detected is an anti-elderly bias. In elderly people show this bias to the same extent as young people do. That’s how strong it is in just about every culture, even though cultures do vary in how much age bias they show, every culture is in that direction of elderly bad, young and good. Yeah. And that’s not changing over time. And we have to ask why. You know, and I have a feeling that. I mean, I can give many it’s a very complicated set of answers, which are all hypotheses at this time, but I think we’re not working on that bias. Our culture cares about race, we talk about it, we argue about it, we get into fits about it. We talk about sexuality, we pass laws, you know, we do things. But on age bias, or disability bias, or body weight bias, those three are not changing at all. In fact, body weight bias, a liking for thin over fat is actually getting worse, over the same 10-year period. Not better.
Jen 50:50
So some good progress, and still a lot of work to do as well.
Dr. Mahzarin Banaji 50:54
Exactly.
Jen 50:55
Alright. Well, thank you so much for spending your time with us today. I’m so grateful. I know how busy you are. And I’m really grateful for the opportunity to read papers before they are publicly available and get a preview into what’s coming and that you were so generous here with your time today.
Dr. Mahzarin Banaji 51:10
Thank you for doing what you do. I think this sort of translation is so necessary, and so few people do it. And so few people do it as well as you so thank you.
Jen 51:20
Well, thank you very much.
Jen 51:21
And so our listeners can find links to all of the references and there are many of them to the studies we’ve discussed, as well as the background reading that I did. And also the book that Dr. Banaji coauthored with Dr. Anthony Greenwald, which is called Blindspot: Hidden Biases of Good People all of that can be found at YourParentingMojo.com/ImplicitBias.
Jen 51:40
Thanks for joining us for this episode of Your Parenting Mojo. Don’t forget to subscribe to the show at YourParentingMojo.com to receive new episode notifications and the FREE guide to seven parenting myths that we can leave behind and join the your parenting Mojo Facebook group for more respectful research based ideas to help kids thrive and make parenting easier for you. I’ll see you next time on Your Parenting Mojo
References
Banaji, M.R., & Greenwald, A.G. (2002). Blindspot: Hidden biases of good people. New York: Delacorte.
Blanton, H., & Jaccard, J. (2008). Unconscious racism: A concept in pursuit of a measure? Annual Review of Sociology 34, 277-297.
Blanton, H., Jaccard, J., Strauts, E., Mitchell, G., & Tetlock, P.E. (2015). Toward a meaningful metric of implicit prejudice. Journal of Applied Psychology 100(5), 1468-1481.
Brown, E.L., Vesely, C.K., & Dallman, L. (2016). Unpacking biases: Developing cultural humility in early childhood and elementary teacher candidates. Teacher Educators’ Journal 9, 75-96.
Cao, J., Kleiman-Weiner, M., & Banaji, M.R. (2017). Statistically inaccurate and morally unfair judgements via base rate intrusion. Nature Human Behavior 1(1), 738-742.
Carlsson, R. & Agerstrom, J. (2016). A closer look at the discrimination outcomes on the IAT Literature. Scandanavian Journal of Psychology 57, 278-287.
Charlesworth, T.E.S., Kurdi, B., & Banaji, M.R. (2019). Children’s implicit attitude acquisition: Evaluative statements succeed, repeated pairings fail. Developmental Science 23(3), e12911.
Charlesworth, T.E.S., Hudson, S.T.J., Cogsdill, E.J., Spelke, E.S., & Banaji, M.R. (2019). Children use targets’ facial appearance to guide and predict social behavior. Developmental Psychology 55(7), 1400.
Charlesworth, T.E.S., & Banaji, M. (2019). Patterns of implicit and explicit attitudes: I. Long-term change and stability from 2007-2016. Psychological Science 30(2), 174-192.
Chugh, D. (2004). Societal and managerial implications of implicit social cognition: Why milliseconds matter. Social Justice Research 17(2), 203-222.
Cvencek, D., Meltzoff, A. N., Maddox, C. D., Nosek, B. A., Rudman, L. A., Devos, T. Dunham, Y., Baron, A. S., Steffens, M. C., Lane, K., Horcajo, J., Ashburn-Nardo, L., Quinby, A., Srivastava, S. B., Schmidt, K., Aidman, E., Tang, E., Farnham, S., Mellott, D. S., Banaji, M. R., & Greenwald, A. G. (in press). Meta-analytic use of Balanced Identity Theory to validate the Implicit Association Test. Personality and Social Psychology Bulletin.
Forscher, P.S., Lai, C.K., Axt, J.R., Ebersole, C.R., Herman, M., Devine, P.G., & Nosek, B.A. (2019). A meta-analysis of procedures to change implicit measures.
Gawronski, B., & Bodenhausen, G.V. (2017). Beyond persons and situations: An interactionist approach to understanding implicit bias. Psychological Inquiry 28(4), 268-272.
Goode, E. (1998). A computer diagnosis of prejudice. The New York Times. Retrieved from https://www.nytimes.com/1998/10/13/health/a-computer-diagnosis-of-prejudice.html
Greenwald, A.G., & Lai, C.K. (2020). Implicit social cognition. Annual Review of Psychology 71, 419-445.
Greenwald, A.G., Banaji, M.R., & Nosek, B.A. (2015). Statistically small effects of the Implicit Association Test can have societally large effects. Journal of Personality and Social Psychology 108, 553-561.
Kang, S.K., DeCelles, K.A., Tilcsik, A., & Jun, S. (2016). Whitened resumes: Race and self-precentation in the labor market. Administrative Science Quarterly 1-34.
Kurdi, B., Gershman, S.J., & Banaji, M.R. (2019). Model-free and model-based learning processes in the updating of explicit and implicit evaluations. Proceedings of the National Academy of Sciences 116(13), 6035-6044.
Kurdi, B., Mann, T.C., Charlesworth, T.E.S., & Banaji, M.R. (2019). The relationship between implicit intergroup attitudes and beliefs. Proceedings of the National Academy of Sciences 116(13), 5862-5871.
Kurdi, B., & Banaji, M.R. (2019). Attitude change via repeated evaluative pairings versus evaluative statements: Shared and unique features. Journal of Personality and Social Psychology 116(5), 681.
Kurdi, B., Seitchik, A.E., Axt, J.R., Carroll, T.J., Karapetyan, A., Kaushik, N., Tomezsko, D., Greenwald, A.G., & Banaji, M.R. (2019). Relationship between the Implicit Association Test and intergroup behavior: A meta-analysis. American Psychologist 74(5), 569.
Leech, T.G.J., Irby-Shasanmi, A., & Mitchell, A.L. (2019). “Are you accepting new patients?” A pilot field experiment on telephone-based gatekeeping and Black patients’ access to pediatric care. Health Services Research 54, 234-242.
Mann, T.C., Kurdi, B., & Banaji, M.R. (2019). How effectively can implicit evaluations be updated? Using evaluative statements after aversive repeated evaluative pairings. Journal of Experimental Psychology: General 149(6), 1169.
Mitchell, G., & Tetlock, P.E. (2017). Popularity as a poor proxy for utility: The case of implicit prejudice. In S.O. Lillenfeld & I.D. Waldman (Eds), Psychological Science Under Scrutiny: Recent Challenges and Proposed Solutions (p.164-195). Hoboken: Wiley-Blackwell.
Oswald, F.L., Mitchell, G., Blanton, H., Jaccard, J., & Tetlock, P.E. (2013). Predicting ethnic and racial discrimination: A meta-analysis of IAT criterion studies. Journal of Personality and Social Psychology 105(2), 171-192.
Oswald, F.L., Mitchell, G., Blanton, H., Jaccard, J., & Tetlock, P.E. (2015). Using the IAT to predict ethnic and racial discrimination: Small effect sizes of unknown societal significance. Journal of Personality and Social Psychology 108(4), 562-571.
Papageorge, N.W., Gershenson, S., & Kang, K. (2016). Teacher expectations matter. IZA Discussion Papers No. 10165, Institute for the Study of Labor (IZA), Bonn.
Payne, B.K., Vuletich, H.A., & Lundberg, K.B. (2017). The bias of crowds: How implicit bias bridges personal and systemic prejudice. Psychological Inquiry 28(4), 233-248.
Rae, J.R., Newheiser, A-K., & Olson, K.R. (2015). Exposure to racial out-groups and implicit race bias in the United States. Social Psychological and Personality Science 6(5), 535-543.
Reyes, R.M., Thompson, W., & Bower, G. (1980). Judgemental biases resulting from differing availabilities of arguments. Journal of Personality and Social Psychology 39(1), 2-12.
Singal, J. (2017, January 11). Psychology’s favorite tool for measuring racism isn’t up to the job. The Cut. Retrieved from https://www.thecut.com/2017/01/psychologys-racism-measuring-tool-isnt-up-to-the-job.html
Staats, C., Captosto, K., Tenney, L., & Mamo, S. (2017). State of the science: Implicit bias review. Kirwan Institute for the Study of Race and Ethnicity, The Ohio State University.
