Implicit Association Tests

A trio of psychology professors from UVA, Harvard, and UW (Washington, nice try Wisconsinites) collaborated on a study investigating the unconscious preferences and beliefs that people manage to shield from others and themselves. While originally designed to plumb the mental recesses of research subjects to determine the roots of thinking and feeling, the researchers have adopted their Implicit Association Test (IAT) for joe shmoe’s like you and me to “gain greater awareness about [our] own unconscious preferences and beliefs.” The test purportedly seeks to root-out both attitudes that we intentionally shield from others, and those that we’re oblivious about to begin with. It’s an interesting approach and I highly recommend trying a couple tests to a) see if you’re a racist/sexist/[insert other rancorous-ist term] bastard and, b) determine whether you agree with their approach.

I think it’s way better if you take the tests in the silken scarves of a virginal IAT subject, so I’ll spare you the bulk of my analysis. Quickly, however, I’m interested if others have qualms with any of the following:

• Many of the tests are structured so that you’re more prone to err in sections indicating traditional biases (based on the precept that you become more adept and comfortable with the test as you answer more questions…I suppose the counterargument would say you become tired near the end and make more mistakes. I’d reply by saying the tests are too short to warrant such an argument…and then we’d go on and on like baboons drunk on our own pseudo-intellectual gibberish)
• Some technical impediments (poor screen resolution, slow computer) could skew results
• Semantics play a pivotal role and I’m not sure the loaded term ‘preference’ can properly encapsulate all the sundry mentalities that score similarly
• Two simple metrics-speed and accuracy-seem a bit insufficient

I’m guessing those who devised the test would either disagree with me or say that nothing’s perfect. I would, in turn, agree with them that I don’t know what I’m talking about and concede that they’re also correct on their second point. In all seriousness, I find this subject fascinating and applaud them for taking steps to investigate hidden biases, because I know I harbor them and i'm guessing others might, too. It's the first step in solving a larger problem. So, without further ado, go try a test because they’re pretty interesting and maybe somewhat revealing.