Seems I'm the Least Racist White Woman in America
I know I shouldn't have, but it was irresistible. I tried the Project Implicit test, the one that's supposed to reveal your secret racial preferences. It seems I have them, but not in the way you'd guess:
***
You have completed the Light Skin - Dark Skin IAT.
Your Result:
Your data suggest a strong automatic preference for Dark Skin compared to Light Skin.
Thank you for your participation. Just below is a breakdown of the scores generated by others. Most respondents find it easier to associate Dark Skin with Bad and Light Skin with Good compared to the reverse.
Huh. How do you explain this? Do you think it's just that I've been living for a long time in a country where most people are more dark-skinned than me? Or is it something else?
I sure didn't think I was in any way unusual like this.
I guess I'm glad that I have scientific proof that I'm not a racist, but something in me says the test is flawed--can't put my finger on what it is--can you?
- Comment (54)
- · Quote
- · UnfollowFollow (2)
Comments:
Oct '10
Re: Seems I'm the Least Racist White Woman in America
They said I exhibited a slight preference for Obama over Santorum. That could not be more absurd. (What's the saying here about a syphilitic camel?)
Clearly, the test has zero credibility, at least in my eyes.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
I just did an IAT for Race. I got a slight automatic preference for African-American compared to white. It seems that that is rare. I also, earlier, got a slight automatic preference for lighter skin over darker skin.
If pushed really hard and trying to be reallly honest, I'd have guessed that I had a preference for darker skin and a slight preference for white over African-American. (I'm a pale white person with a gorgeous Asian adopted child--thus preference for darker skin--and I spend relatively little time around African-Americans, so I'd have guessed that African-American faces would be less familiar and appealing.) Seeing each person as an individual is important to me, though; you'd have to push me hard to get me to confess these guesses.
Nevertheless, I'd be really interested in Aodhan's further explanations about the test and racial preferences, since many of us were surprised by our results.
Aug '10
Re: Seems I'm the Least Racist White Woman in America
You know you're a redneck......if yore danged ipad wont download a flash player to do the goldarned test !
Re: Seems I'm the Least Racist White Woman in America
I tried a few more of those and had results I wouldn't have predicted on them all--it seems I believe other people to be unhappy and anxious, too. (I strongly identified "me" with "happy," and "them" with "miserable.") This test was supposed to measure latent depression, but you could also say it measures latent misanthropy.
But I suspect part of what's being measured here is your skill at the game. How quickly can you grasp that "getting it right" means quickly shifting the mental association when they tell you to? How flexible are you at doing that? I seem to be able to learn to pair words and images--"dark=good," "dark=bad"--quite quickly, and to reverse those pairings just as quickly. The difference between the game and life is that the game asks you to do it and rewards you immediately for doing it correctly.
I wonder if people should start with a baseline test--associating red squares and yellow squares with positive and negative words--to evaluate their skill at the game, first. Also wonder how much slight change in facial expressions change the result--would slightly-smiling dark faces generate substantially different results?
May '11
Re: Seems I'm the Least Racist White Woman in America
I had trouble on the test categorizing some of the faces properly. Many of the "white" subjects in the photos looked like mixed-race, fair-skinned, " black" people whom I've known.
May '10
Re: Seems I'm the Least Racist White Woman in America
For what it's worth, I don't think this test works for people who play video games and have thus honed their ability to match mental agility to finger action. I believed the test when it said I had no racial bias, but I got a pretty good chuckle when it said I felt equally with regard to Ron Paul and Barak Obama.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
Hi Claire,
Your suspicions are partly correct. Skill and speed do confound IAT results.
This is shown, for example, by the fact that people's scores on thematically unrelated IATs are correlated.
However, the confound is partial, not total. Furthermore, the magnitude of the confound can be reduced by statistical adjustment.
.
Claire Berlinski, Ed.:
But I suspect part of what's being measured here is your skill at the game. How quickly can you grasp that "getting it right" means quickly shifting the mental association when they tell you to? How flexible are you at doing that? I seem to be able to learn to pair words and images--"dark=good," "dark=bad"--quite quickly, and to reverse those pairings just as quickly. The difference between the game and life is that the gameasksyou to do it and rewards you immediately for doing it correctly.
I wonder if people should start with a baseline test--associating red squares and yellow squares with positive and negative words--to evaluate their skill at the game, first. Also wonder how much slight change in facial expressions change the result--would slightly-smiling dark faces generate substantially different results? · 6 hours ago
Nov '10
Re: Seems I'm the Least Racist White Woman in America
I would note that in the Gawker interview, John Derbyshire challenges the NR people to make their IAT results public, apparently in confident expectation that everyone is as racist as he is. I am glad that at least on Ricochet his expectations were not met.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
You may be right .There is some evidence that IAT effects are smaller for people who have faster reaction times, such as younger people. This may be tthe case, not only when the faster times are due to maturational effects, but also when they are due to practice.
Note also that the authors of the test claim is that implicit attitudes can unexpectedly diverge from explicit ones. If so, then noting such a unexpected divergence, as you do, does not logically invalidate the claim, but is rather consistent with it.
Of course, a divergence does not establish that the claim is true; further evidence would be required. One type would be predictive validity. The IAT has some of that.
.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
You are entirely correct.
Interpreting what IAT results mean can be very tricky.
Here's another example. Take an IAT with the categories Young, Old, Good, and Bad. Many people go faster in the Y+G/O+B block than in the Y+B/O+G block. This has been interpreted as demonstrating implicit ageism: a relative prejudice against old people in favour of young people.
However, no less plausible is the following interpretation: an abstract preference for youth instead of age. While not everything about growing old is to be decried, a preference for youth is perfectly defensible and natural.
I would certainly prefer my mind to be implanted in a younger body, were that possible. I anticipate this preference intensifying as my senescence proceeds.
Si jeunesse savait, si vieillesse pouvait...
.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
Your secret passion has been unearthed for all to see...
.
dittoheadadt: They said I exhibited a slight preference for Obama over Santorum. That could not be more absurd. (What's the saying here about a syphilitic camel?)
Clearly, the test has zero credibility, at least in my eyes. · 10 hours ago
Nov '10
Re: Seems I'm the Least Racist White Woman in America
This could be true, but it isn't, and is a common legitimate suspicion on first encountering the test. See some other posts on the subject.
.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
An excellent suggestion. Here's a slight elaboration.
Suppose you had four IATs, with the categories below.
(Abbreviations: s = smiling / f = frowning / b = black / w = white)
(1) Whites (s), Blacks (f), Positive, Negative
(2) Whites (f), Blacks (s), Positive, Negative
(3) Smiling (w), Frowning (b), Positive, Negative
(4) Smiling (b), Frowning (w), Positive, Negative
One could check whether (by comparing 1 to 2) facial expression moderated the Race IAT effect, and whether (by comparing 3 to 4) race moderated the Facial Expression IAT effect.
I would predict, intuitively, that an attenuation but not a reversal of the Race IAT effect in 2 compared to 1, but no difference in the Frowning IAT effect for 3 and 4.
If so, the interpretation would not be straightforward. But my reading of it would be that implicit frownism would have been shown to be worse than implicit racism.
I could run the study. Would anyone like to sponsor it?
.
Nov '10
Re: Seems I'm the Least Racist White Woman in America
Bad test. All classifications should be unambiguous.
.