More on Implicit Association Tests
01/18/18 11:35
The Implicit Association Test came up in lecture on Tuesday as we discussed how to determine racism (or other bigoted beliefs). Jennifer, one of our wonderful GSIs, shared new information about how the test might not actually be so effective. She sent this along:
I found this article that articulates the main criticisms of the IAT:
https://qz.com/1144504/the-world-is-relying-on-a-flawed-psychological-test-to-fight-racism/
She discussed that the IAT only weakly predicts behavior (and not more so than explicit tests of racism), the reliability is weak (0.5 at the high end which is not acceptable in psychology), people can predict their own IAT results suggesting that it's not actually unconscious or something that people are unaware of, trainings that use the IAT are not effective in changing behavior, and so on. It's a bit long, but I think she does a good job at describing why it's flawed for both academics and non-academics while recognizing that it could be useful in some ways.
I found this article that articulates the main criticisms of the IAT:
https://qz.com/1144504/the-world-is-relying-on-a-flawed-psychological-test-to-fight-racism/
She discussed that the IAT only weakly predicts behavior (and not more so than explicit tests of racism), the reliability is weak (0.5 at the high end which is not acceptable in psychology), people can predict their own IAT results suggesting that it's not actually unconscious or something that people are unaware of, trainings that use the IAT are not effective in changing behavior, and so on. It's a bit long, but I think she does a good job at describing why it's flawed for both academics and non-academics while recognizing that it could be useful in some ways.