The Washington Post magazine had an interesting article (sorry, registration required) about an instrument that purports to measure racial and gender bias. A group of researchers at Harvard have designed it, and a version is available on-line. [Edit: It looks like the site is getting slammed. Patience, grasshopper.]

In a nutshell, the test goes as follows: round 1 is a series of faces, which you flag “black” or “white”. They’re a series of plain yearbook photos. Round 2 is a series of words that you flag as “good” (by pressing the one key) or “bad” (by pressing another). Good words include love, beauty, joy; bad words include evil, failure, and nasty. I think the first two rounds are mainly to familiarize you with the words and images.

Then the fun begins. Future rounds of the test have you categorize the faces and words into “Black OR Good” on one hand, and “White OR Bad” on the other. In another round you might do “Black OR Bad” and “White OR Good.” These categories also switch sides a couple of times to help balance a preference for one hand over another. The whole thing takes about 10 minutes, and you are instructed to move quickly through.

The contention is that, in people with a bias towards whites for example, they are more likely to trip up and put a “Good” quality in the “White” column and a “Bad” one in the “Black” column, even if that particular round is asking you to put Black and Good items in the same column. Or the test-takers may sort the terms appropriately, but it will take them longer to figure out the correct column, and that counts against them as well.

According to the article, 87% of whites exhibit some level of bias against blacks, and almost 50% of blacks share that bias. Many of the researchers themselves were chastened to find themselves in that 87%. And interestingly, people who self-identify as liberal have a lower instance of bias than people who self-identify conservative. (Those self-designations have no bearing on the scoring of the test, as far as I can see.)

The researchers who put all this together are quick to say that attitudes do not necessarily translate to action. An instrument like this doesn’t necessarily predict racist or sexist behavior, just split-second perceptions—which certainly does put Malcolm Gladwell's most recent book, Blink: The Power of Thinking Without Thinking, in an interesting light. I haven't read the book yet, but my understanding is that the book deals with a cognitive activity he calls "thin-slicing," which is our tendency to make quick judgments based on rapid assimilation of information. Gladwell argues that much of the time, those quick judgments are correct, or correct enough. As someone interested in spiritual matters, I am wondering what place contemplation and discernment play in a world of snap judgments, but that is a topic for another time.

So, of course you want to know, Gentle Reader—did the RM household take this test? You better believe it. Mr. RM took it first, and exhibited no bias. I, on the other hand, exhibited a slight bias—towards black people.

Yes, black people. That’s right, I’m sick of blacks in this country being hassled by The Man, even though I am The Man.

Speaking of the man, the gender test seemed trickier to me. The categories were Male/Female (using names rather than faces this time), and Family/Career. Family words included marriage, wedding, and children; and career words included business, management, and office. I exhibited a slight tendency towards correlating males with career concepts, and Mr. RM was strongly likely to correlate those concepts.

So he’ll be sleeping on the couch tonight.

But seriously, this is a man who cooks all our meals, who never describes time alone with his child as “babysitting,” whose mother campaigned for the ERA, and who routinely glanced through my Ms. Magazine before I even had a chance to. So, we don’t know what’s up with that, but we’ve had fun talking about the issues the test raises, as well as its possible limitations.

Anyway, check out the tests, as well as the article. The article talks about other studies that have been done about the issue of bias, including one in which a fictional Canadian man convicted of a crime receives a shorter sentence from a mock jury than a fictional Mexican man convicted of the same crime. Another experiment involved sending out resumes to various companies, which showed that a moderately qualified man with a white-sounding name like Greg received more interest than a well-qualified man who happened to be named Tyrone.

It’s a very interesting topic, and regardless of one's feelings about the test, it's an opportunity for some soul-searching, which is a good thing in my book. And it can open up some fascinating dialogue, probably more so than Which Peanuts Character Are You? which is admittedly entertaining in its own right.

Signed, Linus.