In an earlier post, I described Malcolm Gladwell’s book Blink about the way we instinctively make judgments about people. The way we make snap judgments is by ‘thin-slicing’ events. We take in a small slice of the phenomena we observe and associate the information in those slices with other measures. People who make good snap judgments are those people who associate the thin-slice information with valid predictors of behavior. People who make poor or prejudicial judgments are those people who associate the thin-slice information with poor predictors.

Think about what you observe about a person immediately as that person walks into your view. Gender, ethnicity, height, weight, color, gait, dress, hair, demeanor, eyes, looks, physique, gestures, voice, the list just goes on. We sweep up all these impressions in a flash. And based on them, whether we want to or not, we make a judgment about the person. Different people will weigh different elements in the mix differently.

If someone comes into my office wearing a suit, my initial impression of the person is different than if she had come in wearing jeans. (If you were mildly surprised by my using the pronoun ‘she’ towards the end of the last sentence, it is because, like me, you implicitly associate suits with male attire, so that the first part of the sentence made you conjure up a mental image of a man.)

A personal example of snap judgments occurs when I read Physics Today which I get every month. The obituary notices in have the magazine have a standard form. There is a head-shot of the person, with the name as the header, and one or two column inches describing the person.

Almost all of the obituaries are of old white men, not surprising for physicists of the generation that is now passing away. I found myself looking at the photo and immediately identifying whether the person was of English nationality or not. And I was right a surprising number of times. And I was not reasoning it through in any conscious way. As soon as I saw the picture came into view, I’d find myself thinking “English” or “not English”. I don’t know the basis of my judgments. But as I said, I was right surprisingly often.

Gladwell describes a very successful car salesman who over the years has realized that gender, ethnicity, clothes, etc. are not good predictors of whether the person is likely to buy a car or not. Someone who his fellow salespeople might ignore or dismiss because he looks like a rustic farmer, this salesman takes seriously. And because this salesman has been able to shape his intuition to ignore superficial or irrelevant things, his senses are better attuned to pick up on those cues that really matter.

Some of the strongest associations we make are those based on ethnicity, gender, and age. We immediately associate those qualities with generalizations associated with those groupings.

People are not always comfortable talking about their attitudes on race, gender, and other controversial topics. This is why surveys on such topics are unreliable, because people can ‘psyche out’ the tests, answering in the way they think they are expected to, the ‘correct’ way, rather than what they actually feel. This is why opinion polls on such matters, or in elections where the candidates are of different races or ethnicities, are hard to rely on.

There is a website, developed by researchers at Harvard University, that recognizes this problem. They have designed a survey instrument that tries to overcome this feature by essentially (as far as I can tell) measuring the time taken to answer their questions. In other words, they are measuring the time taken for you to psyche out the test. Since we have much less control over this, the researchers believe that this survey gives a better result. They claim that you cannot change your score by simply taking the test over and over again and becoming familiar with it.

If you want to check it out for yourself, go to the test site, click on “Demonstration”, then on “Go to Demonstration Tests”, then on “I wish to proceed”. This takes you to a list of Implicit Association Tests (or IAT) and you can choose which kinds of associations you wish to check that you make.

I took the Race IAT because that was what was discussed in Gladwell’s book, and it took me less than five minutes to complete. This test looks at the role that race plays in making associations. In particular it looks at whether we instinctively associate black/white people with good/bad qualities.

It turns out that more than 80% of people who have taken this test have pro-white associations, meaning that they tend to associate good qualities with white people and bad qualities with black people. This does not mean that such people are racists. They may well be very opposed to any kind of racist thinking or policies. What these tests are measuring are unconscious associations that we pick up (from the media, the people we know, our community, etc.) without being aware of them, that we have little control over.

Gladwell himself says that the test “always leaves me feeling a bit creepy.” He found himself being rated as having a moderate automatic preference for whites although he labels himself half black because his mother is Jamaican.

I can see why this kind of test is unnerving. It may shake our image of ourselves and reveal to us the presence of prejudices that we wish we did not have. But if we are unconsciously making associations of whatever kind, isn’t it better to know this so that we can take steps to correct for them if necessary? The successful car salesman became so because he realized that people in his profession made a lot of the unconscious associations that were not valid and had to be rejected. And he used that knowledge in ways that benefited him and his customers.

Although you cannot change your Race IAT scores by simply redoing the test, there are other things that can change your score. When I took the Race IAT, the results indicated that I have no automatic preference for blacks or whites. In a later posting, I will talk about the effects that ‘priming’ might have on the test results, and how that might have affected my results.