Then there are moral problems with stereotypes. So, let's assume that the stereotypes we possess are accurate. Still, even so, even if they're statistically grounded, we often have the intuition that people should be judged as individuals, not as group members. Now, this gets complicated for whatever reason society allows certain circumstances where you can judge people as group members. In the United States, for instance, auto insurance, how much money you pay to insure your driving and your car, varies depending on how old you are and other facts about you, like even your grades if you're a teenager. The auto insurers would ground this data that, for instance, younger drivers are more at risk than middle-aged drivers. But at the same time, you wonder whether that would be moral if, for instance, different ethnicities had different records of safety in driving. Similarly, we allow some profiling in criminal cases. For instance, there's a strong presupposition when there's a violent crime that the perpetrator is male just because there's so much evidence that men are much more prone to human violent crimes than women. But it gets complicated again to look at ethnicities, and one might say reasonably enough, and this is a philosophical normative debate, not a psychological one, that even if you gain some benefit through racial profiling, the cost in terms of making it to life worse for minority members of the society outweighs any benefit that you get. Another further practical issue is that when you're dealing with terrorists groups or other intelligent and benevolent people, if you rely too much on the category, they will change the category in order to thwart you. That is, if you're looking only for males of a certain age and certain ethnicity because they are statistically connected to the criminal or terrorist organization you're looking at, the people in criminal or terrorist organization are not stupid, and they'll act before you. They'll act, for instance, to recruit people who don't look like they're members of that organization. Social psychologists are interested in all social aspects of stereotypes and they tend to talk about different levels and different types of stereotypes, and you can imagine a distinction between three levels. One is public stereotypes, which is what we say to other people about a group, the ones we're just upfront about. I'll say men are larger than women. I believe that, on average, men are larger than women. A lot of women are larger than men but on the most part blah blah blah. That's a public stereotype. But then there's private stereotype, which I really believe to be true but I'd never say it to you, not merely in a lecture. I just won't say it to anybody, I just know when I say, "I shouldn't say this." Then, and this is perhaps the most interesting case, there's implicit stereotypes, which are unconscious associations that guide us and affect us even if we're not conscious of it. Now, we know from a lot of work that public stereotypes have been changing. These are examples from the United States. So, for instance, if you ask people how likely they are to vote for a qualified African-American to be President, if you ask them in '50s and '60s, about half people would say, "No, I won't do it. I don't want an African-American as president." Now if you ask now, just about everybody says yes, and actual voting patterns for Barrack Obama indicated even people who are entirely white communities are quite willing to vote for an African-American to be president. Some will ask people with different attitudes in this case about blacks. People ascribe them horrible traits. But as time goes by, they're just less likely to do so, and by 2000 last time, the study was done just about nobody does as is the change in the public stereotypes. The change in the private stereotypes is actually a lot harder to look at almost by definition. So we know a lot less about this. There's actually some very exciting work which is in progress, so I don't have that much to talk about looking at Google searches. So, it's to a surprising extent and I think a shockingly disappointing instance, for instance, people use ethnic slurs and racial attacks and a lot of expression of racial hatred when they are in private typing things into their computer. This would suggest that the're a lot of great changes in the public stereotypes. Private stereotypes have been slower to change. Then there's implicit attitudes. This is a topic of tremendous focus and tremendous interest by social psychologists. So, for instance, in one study, what you might do is you might flash pictures of black faces and white faces subliminally on a computer screen so fast that people don't even know what they're seeing. They don't know they're seeing a face. Then you give people words to complete like hos. In this study, they are more likely to complete a word like hos with a word like hostile, if seeing the black face, but hospital if seeing a white face. This is a suggestion that even below the level of consciousness, we have different attitudes towards different racial groups. I'll add that these effects show up not just for white people but also for black people. That is, a lot of implicit biases show up in all groups including the group that's biased against. One exciting method to explore this is the so-called IAT method. In fact, if you go implicit.harvard.edu, you could do this yourself. These are basically reaction time methods looking at how quick you are to associate the positive with some groups and a negative of other groups. Here's how the game is played. You'll see images. If there is something that is white-American or good, press it with your left hand. If it is African-American or bad, either a face or the word, press it with your right hand. Try to do as quick as possible, like this, that would be good. That's bad right, that's a white face left, that's an African-American face, right. So we get your speed at doing this, but then we test you on opposite thing where you connect up African-American are good, and white American are bad, and then we show you different words and different pictures. The logic here is, if you naturally have an association, say between white-American and good, you'll be more fluid in making the responses when the two words are put together than when they're distinguished. Now focusing on race here, with IAT group has an enormous range of studies of different biases. We look at politics and associations between different skills and different genders. Things like religion, things like age, one of the strongest biases, I hate to say this one, the strongest biases people have is against old people. There is little to be said for old people at least implicitly. This is, as I said, really interesting. It's really interesting because it tells us a lot about psychology and a lot about how the mind works but also as real-world relevance. You might say, "Look, this is just little studies, differences of a fraction of a second." For one thing, fractions of a second can really matter. There's evidence, for instance, that there are police officers and also people in simulations are more likely to shoot a black person holding a harmless object than a white person. The decision whether to shoot or not is made in fraction of a second, and these biases can apply. Also, we often face judgment calls. These are some typical studies, for instance, which find that if you're the only witness and somebody is in trouble, you're just as likely to help them if they're white or black. But if there are other people and you're not sure maybe someone else will help, do I really have to help? You're more likely to help if the person is white than black, and this is if you yourself are white, suggesting a racist bias you may not even be aware of. Here's some other work. I love to say a recent study that went on auction site eBay, and it turns out people will bid more for these cards than for these cards even though cards are identical. It's just in the hands that hold them; white versus black are different. The people who are affected by the color of the hands may be unaware they are affected. It is an implicit subtle level. It changes your attitudes and your feelings. We're at war for ourselves then. We're not at war of ourselves if you're a racist and proud of it and want to distinguish between races and have strong racist views, you are if anything consistent. Many of the rest of us are at war for ourselves in that group membership matters even when we think it shouldn't. Even when we think we want to be colorblind, we want to be gender-blind, we want to be attractiveness blind, we're still swayed by things. It's an interesting question how to deal with this problem and I think part of the answer is, we can use our intelligence to override our biases. That is, we don't become less biased just by trying to be less bias rather we become less bias by using clever methods to override it. So, for instance, it used to be that for auditions for symphony orchestras, men would be overwhelmingly chosen over women, and because they sounded better according to the judges but it turns out that when you do blind auditions, you have them audition behind the screen, the effect goes away, and a gender difference largely disappears. What's interesting here from a psychological point of view is the judges, male and female who were saying that the men sounded better than the women, were most likely not being self-consciously sexist saying," Ha ha I'm going to discriminate against women. I hate women," etc. Rather they might have entirely had egalitarian views, but because of their biases, they couldn't help but hear them differently. Then when you hide the gender of people, you take away the chance for the biases to operate. Things change. More generally, we talked about so many social psychology biases, so many ways in which we're contorted by first impressions, by stereotypes that we believe are inaccurate, by cognitive dissonance. The way we can make it through the world as better decision-makers and better people often is to use our intelligence to structure the world so these biases don't apply. To put it much broader, we can use our heads to override our hearts.