Libertarian Librarian

Or, to quote a Mr. Bill Shakespeare, "Told by an idiot, full of sound and fury, signifying nothing."

Saw a piece on my local news last evening about a new way of determining if you're racist. It's called an Implicit Association Test (IAT), and it's all the rage at Harvard, where it was developed. The rationale for the IAT being a legitimate means of determining your implicit (ie, subconscious or without control) reactions to various concepts or people is that it is easier to associate things you implicitly like with words that are good and harder to associate things you implicitly like with words that are bad. The test determines which words you associate with what based on how long it takes you to respond. Okay, if that makes no sense go here-- that should clear it up.

Anyway, in theory the test measures how quickly you tend to associate a particular group (say black people) with good words or bad words relative to how quickly you associate a different group (say white people) with good words and bad words. Thus revealing if you subconsiously prefer one group over the other.

In theory. In practice, I found myself trying to anticipate what the next picture or word would be so that I could answer faster. "Five good words in a row, the next one has got to be bad," would be an example of my thinking during much of the test. Now, maybe I'm a freak (shaddup!), but I fail to really see how the test measured anything other than my ability to take the test.

My believe that the test is pretty much crap is supported by the results. I took the race test twice, getting "slight implicity preference for Eurpoean Americans over African Americans" the first time, and no discernible difference the second. I took the Arab test once (and the Arab test is especially stupid since it just uses names, no pictures, and the non-arab names are so strange as to make implicit associations with anything unlikely in my opinion) getting the slight preference for other people over Arabs.

Which is believable. I'd like to believe I have no preconceived stereotypes or "implicit" reactions to people based on color/nationality, but I can buy that I have a small knee-jerk preference for folks who look more like me. But then I take the Presidents IAT.

Yikes.

First test, I had no preference between FDR and G.W. Bush. Second test, I had no preference between Clinton and G.W. Bush. Third test I had a moderate (not slight) preference for G.W. Bush OVER Ronald Reagan.

Puh... leeze. First of all, I'd take FDR over Bush without a qualm, so for me not to have an implicit attitude seems unlikely. But secondly, I have no preconceived or well-considered conceptions of Bush being a better president than Reagan. I mean, I have Reagan listed at #4 on my top 10 presidents list. Bush would be in the bottom half, and possibly the bottom quarter of that list.

Finally, I take the Barry Bonds vs. Babe Ruth test. So-- I've already tested to a slight preference for white folks (Ruth) over black folks (Bonds). Plus, I detest Barry Bonds as a Jag and a cheat, while respecting Ruth for being a phenomenal athlete, if not a particularly nice person. Simply put, I like and respect Ruth a lot more than Bonds.

Given that, I would have expected the IAT to come out with at least a slight preference for Ruth over Bonds, and a moderate or even extreme preference wouldn't have seemed that odd when you combine my dislike for Bonds with my ealier implicit preference for white people over black people. Result? Little to no implicit preference between the two.

This test isn't about implicit impressions-- it's about how fast your brain can accurately process data and then get one of your fingers to click a keypad. The tests are also kind of stupid because some of them have pictures and others just have words, thus "testing" two different parts of the brain, and therefore adding another variable into the results.

I wonder how much time and money has been spent on this pointless test (it's ongoing-- so someone is buying it)? And I wonder how a news broadcast covered this test as if it were a legitimate means of determining if you are a racist or not (that's how they billed the 2 minute piece they ran)?

Oh, wait. It's the mainstream media. Never mind about that second question.

Labels: Oh That Ivory Tower