I'm Mahzarin Banaji. I'm an experimental psychologist. I teach at Harvard University. I've been fascinated by two things. We call them attitudes and we call them beliefs. These are thoughts and feelings that are in our minds. We can't see them, we can't touch them, so how to go about measuring them is the interesting question. For more than a 100 years, psychologists have been at this task of trying to understand, trying to get out of people what their thoughts and feelings are. These particular thoughts and feelings that I'm interested in, are attitudes and beliefs that we hold about social groups, and how we use those beliefs that we have about the group as a whole: The Swiss are like this, so the Germans are like that. How do we use those beliefs? Then our judgments about individual people who happened to be members of the group. That's the interesting question and as I said, for over a 100 years, people have been measuring them in the form of attitudes or prejudices that we might hold, thoughts, or beliefs, or stereotypes that we may have about different groups of people. How to go about doing this is the question. We know one easy way to do it, and that is, to ask people, what do you think about x? How likely is it that people from this group will be able to do something of this kind or that kind? What is their inner nature? Are they competitive? Are they collaborative? Are they smart or they not very smart? Are they trustworthy? Are they not trustworthy? Things like that, so we would just ask. But then asking has some problems. There are two problems with asking people. First of all, if you ask me what I think about x, I may not want to tell you about that because that's a private belief and I don't want to share it with you. But there's a second problem with asking, and that is the one I'm most interested in. That is, that we may not know our own minds. Our brains do a lot of work silently, quietly. We don't have access to all of it, and we don't necessarily have access to the contents of our mind. We don't know, and this may seem strange to you, but we don't know what our actual attitudes might be or what thoughts and beliefs are. As scientists, our job is to try to figure out indirect ways into people's minds to draw out what might actually exist there that even they may not know. The work that I have been doing for the last, close to 40 years now, is work on trying to get at the implicit, or the unconscious, or the less visible even to ourselves, contents of our minds. We've developed tests, tests like the IAT, the Implicit Association Test, which we use to bypass, to get away from asking people what they might think or feel, and instead, just trying to see what might be in their minds. We might look, for example, at how rapidly, or how quickly, or how accurately you can put two things together. The test will reveal, in a sense, what it is that might be sitting in your mind that you may not know about. I'll give you an example, just with me. I am a woman, I have worked outside the home all my life, and yet when I take a test that requires me to associate female with career and male with home, I can't seem to do that as well as if you gave me the opposite. If you ask me to associate female with home, male with career, that turns out to be relatively easy for my brain to do. Why? I don't have this belief that women don't belong in the workplace or anything like that, and yet my brain contains the thumbprint of the culture in which I live, and that culture has repeatedly associated female and home more so than male and home, and that's now in my head. That's the important discovery that we can say consciously, we can say explicitly, what we think and feel. That's one way in which our minds work. But there is another part of our mind where these associations that are picked up, sucked out of the culture, and sit there, they're there all the time. Where do they come from? One of the things that we've been very interested in, is looking at young children. How do young children come to have the beliefs and attitudes that they do? Do they learn it slowly as they grow up and so on? Our data suggests no, that children are very open to what's going on in their culture, and at a very early age, at ages like two and three, we can see evidence that they have in their minds, attitudes and stereotypes of the sort we see in adults. So what we now know is that these are picked up pretty quickly and that they exist in the minds of children, even at that age. They're not always visible, but we can see them in things like conversations. A student of mine, Tessa Charlesworth, has recently been analyzing data from parents and children. These are conversations that parents and children have had, and there are thousands of such conversations. We can use an approach called natural language processing. We can use machine learning to look at whether in the language that parents are using, they actually are relying on these stereotypes that they have no clue that they are. When you ask parents, they'll tell you, I don't know where my child pick this up from because I certainly don't teach it to them, and yet when we analyze these conversations that parents and children have had, we can see in those conversations evidence that parents are indeed associating, for example, female more with home things and male more with career affairs.