Implicit Attitude Tests

August 28, 2001 9:37 AM

Implicit Attitude Tests are a way to find out how biased you are regarding race, age, and gender. The results can be spooky.
posted by muckster (15 comments total)

Well, I said that male/science and female/arts was strongly environmental and then showed a slight bias in associating the two.

If I was right, in it being environmental, shouldn't I have shown a stronger bias? Or am I not part of the environment?
posted by andrew cooke at 10:10 AM on August 28, 2001

I said I mildly preferred European to African, and it verified that attitude.
I said I strongly associated male with science, and it verified that attitude.

So at least I'm not foolin' myself.

A good tool for self-examination for those who haven't done so honestly.
posted by marknau at 10:27 AM on August 28, 2001

On the black/white Aptitude test, I wonder if they would get the same results if they changed the order so African-American or Good was first instead of last. Would there results be reversed? Does the test really measure what they think it measures?
posted by Apoch at 10:47 AM on August 28, 2001

I think the test changes the order each time a new person does it.
posted by dydecker at 10:54 AM on August 28, 2001

I'm not a psych professional or anything, but this seemed more like a mind conditioning test than anything else. I quickly realized what the test was doing and my head then tried to focus on "beating" the system. Testing whether an indiviual associates negatives with a particular race would certaily prove interesting results, but like Apoch said, I wonder how different the test would be if they initially conditioned the tester's mind and motor-skills that European-Americans == negative as opposed to African-Americans. For what it's worth, the test did provide results that lined up with what I believe.
posted by Hankins at 11:00 AM on August 28, 2001

I wondered about that myself.. But I think they can claim to have 'beaten' that affect by gradually shifting the associations, rather than doing it all at once. They don't just make it "White = Good, Black = Bad" one round, then "Black = Good, White = Bad" the next.. they do it in steps over 4 or 5 rounds, and they eventually have you do the last association twice and probably compare the results of the two rounds to see if you do better with more practice.

I do think that this is a fairly common way of testing, whether it's accurate or not, because it's been part of the only two psychology studies I've ever volunteered for.
posted by Hildago at 11:08 AM on August 28, 2001

Incidentally, my results were not what I expected them to be. I predicted a slight preference for white over black (because everyone knows that white guys are inherently a little racist) and the test results showed no preference either way. Then I predicted a moderate association between men and science, and the test results said I had a slight favoritism towards women and science, which doesn't make any sense to me. And I tried my best to be honest and prompt. What does that mean?
posted by Hildago at 11:13 AM on August 28, 2001

From their FAQ:

1. I had to group White items (faces or names) together with pleasant words first. I then found it difficult when I later had to group Black items with pleasant words. Was that because I had done the task first with White items and pleasant words?

Answer. The order in which tests are administered does make a difference in magnitude of what we call 'IAT effects'. However, it is not a large difference, and only rarely does it change the direction of the outcome. Because of this order effect, the orders used for IATs presented on this website are assigned at random. If you want to check whether the order made a difference for you, you can take the test in the reversed order. If you want to do that now, use your browser Back function to return to the page on which you got your IAT test result. Then click on the option near the bottom of the page labeled Think your results occurred because of the order? Try the other. If you do take the test twice in different orders and get different outcomes, the best estimate of your result is intermediate between the two.


My data suggest little or no preference for White American relative to Black American, suggest a slight identity with Black American relative to White American, and suggests little or no preference for Self relative to Other.

I am a 'White American' male and I found the results interesting. I wonder how much my concerns shaped my results though. My family are what I consider very typical in that they would deny being prejudiced but their actions to the contrary speak far louder then words. I always try to judge the person and not be prejudiced and I have to admit I was more then a bit worried this test would reveal some prejudice I hadn't admitted to myself. Even though it was more reflex and reaction with the flashcards I wonder if my concerns biased my results to any significant degree.
posted by Animus at 11:16 AM on August 28, 2001

Good points about knowing what they are doing and worries about trying to "tool" the test.

I do a number of things routinely that require me to "hide data" from myself and act as if I don't know it. I found myself doing that during the test, and trying to keep a Zen state of just doing the test without any meta-thought.

But, admittedly, I don't know what affect any of my pro-conceptions and meta-processes had on the results.
posted by marknau at 11:30 AM on August 28, 2001

To what Hankins and Apoch mention... I did the age test (results showed no preference either way) and felt I was being trained for young=good. I could do it it again with the test order reversed, but what's interesting and perhaps indicative of implicit prejudice is that my concern was a bias against old.

I thought about some real life scenarios, like if I'm travelling with a huge backpack, and have to pee... who do I ask to watch it for me for a moment, a younger or older person? I ran into a prejudice right there - I would trust an older person over a younger one, for instance.

On the other hand, my sweetie is 8 years younger than me, and to be honest, I always feel a little sheepish telling his friends (most are around his age) how old I am. This implies a negative age bias (aging, specifically).
posted by spandex at 11:56 AM on August 28, 2001

I strongly reccomend a read over their whole faq. Interesting stuff. Well, to those of us social psych geeks anyway.
posted by daver at 3:23 PM on August 28, 2001

My first try resulted in a "slight automatic preference for black." Hmmm. I thought I was fairly unbiased, so I took it again. The second try resulted in a "strong automatic preference for black."

Hmmm. Should I try it a third time?
posted by Poagao at 7:31 PM on August 28, 2001

Thought I'd seen this before... does it count as a double post?
posted by normy at 10:26 PM on August 28, 2001

The gender test was *much* too easy to figure out. I strongly associated women with science, just by noticing what the folks were up to. "His" is placed before "history" and so forth. Bad, bad. Horribly designed li'l test, holes out the wazzoo.
posted by raysmj at 8:01 AM on August 29, 2001

Hrm, it said I had a strong prefrence for whites, but I took the 'white == good' part second. It may have just been practice. This test was actualy pretty difficult for me, I've been tired for a while, and it just confused the hell out of me, mostly.

But still, an intresting test. I'll probably try to take it again when I'm in a better mental state. :P
posted by delmoi at 6:22 PM on August 29, 2001