Implicit Bias

Given the large response the "Political Compass" received, I thought I might throw this other "test" out there for Fantastiko readers to enjoy. It's another one of those self-assessments, but of a less political nature.

In this weekend's "The Washington Post Magazine," they had a very interesting story about prejudice and bias. What I found so interesting is that it was not a story at all concerning the nature and effects of prejudice, but rather its measure; something that I would think very difficult to do since it would in large part, I would guess, depend upon the honesty of whatever survey group they might use. And who can really trust anyone to be honest when it comes to admitting that they have a distaste for fat people or blacks or whites or arabs or gays? Because it's just not cool have an aversion to something different from you.

So a group of psychologists at Harvard University are attempting to measure a person's bias as a function of time. How long does it take you to associate the word "laundry" with "Sarah"? And how long does it take you to associate the word "laundry" with "Tom"? Because if your fingers don't click on Tom/laundry as quickly as they do Sarah/laundry, you hold a negative association with men doing work in the home.

This is the Implicit Association Test and, according to the article,* the results are startling (not really). For example, those who honestly believe that they have no negative bias against gays, find that they actually do--in fact, 38.4% gay respondents have found that they have a negative bias toward gays. A sad figure, but not surprising (since 82.5% straight respondents also hold this negative bias). The article details other findings of course (a token political bend for the sake of the overall Fantastiko theme): conservatives, on average, show higher levels of bias against gays, blacks, and Arabs than liberals; 48.3% of blacks have a bias against blacks.

I took the tests that measure black/white and fat/thin bias. I was skeptical of the method: because you are asked to make associations in a certain order, it seems that the test itself is biased. However, the articles insists (and I found to be true, at least in my case) that the order of things does not matter.

I was surprised by the results of the first test (black/white--I have no negative bias, though I suspected it would tell me I did since I thought I was having the appropriate difficulties taking the test) but not the second (I wrote in the voluntary questionnaire that I do have a slight negative bias toward fat people, and it showed that I did). The article warns that most will be surprised by their results. What it matters, I do not know. Maybe it will turn your world upside down.

--JB

*Sorry I don't have a link to the article--I read it in print and you have to be registered to read it online, which I am not. If you want to check it out, just go to www.washingtonpost.com and search "see no bias"--it's by Shankar Vedantum.