Estimated reading time: 2 minutes, 39 seconds
So as part of the Facing History and Ourselves course I’m taking on News Literacy, I watched this awesome video by Oxford professor/psychologist Binna Kandola on diffusion bias. Watch it 🙂 In addition to being informative, he is also a very engaging speaker.
One of the things he mentions in his talk is something called the Implicit Association Test, which is available here via Harvard (even though it’s not developed by folks at Harvard, which is weird, but… ok).
So if you’ve never heard of this test before, according to the website it “measures attitudes and beliefs that people may be unwilling or unable to report. The IAT may be especially interesting if it shows that you have an implicit attitude that you did not know about.” You get to choose all kinds of different factors/measures.
The first one I tried was women/science. The research question being whether I have an implicit bias about science being a more male field? The answer turned out to be a slight bias. That’s reasonable. I don’t have a strong bias on this, but … socialization and all that. Apparently I hesitated a split-second longer on the questions where I had to put women/science together. You’ll understand what I mean when you try it.
After you finish the test, it asks you some demographic questions (anonymously) which is understandable. It might be useful for them to know how differently men and women respond to the questions of women/science, for example. Now here’s the kicker. After I did the women/science one, I saw an Arab Muslim one and thought hmm this should be interesting, let’s try this. But you know what? They asked the demographic questions BEFORE. And you know what else? They didn’t ask generic questions like your gender and age and religion and ethnicity (all understandable for this particular test), but they delved into a few other things, including where you live, where you’re from, and your postcode. This all posed a red flag for me.
What do YOU think?
I ended up not doing the test altogether. Though I am dying of curiosity.
P.S. after this, I tried the gay/straight one and I got a “slight preference for gay people” (Which I know would not have happened around 3 years ago… this makes me happy … in the sense that knowing people closely can affect our biases … if you wanna believe that test anyway). But I also noticed this test asked for things like postcode, etc., and you can choose to skip the question… so now I’ll re-do the Muslim one while skipping those intrusive questions and see what happens…
P.P.S. Strong automatic preference for Arab/Muslim people. Apparently you can’t hide from this test 🙂 I really did expect myself to have only a slight preference. However, I did not like this particular test because it used only Muslim names, as in proper names, not nouns representing Islam (so not Quran, mosque, Kaaba, but actual names like Wahib and Muhsin – some of the none-Muslim names were not familiar and I wasn’t sure if they were like Pakistani names or non-Muslim altogether so I made some mistakes w that).
3 thoughts on “Implicit Discrimination in the Implicit Association Test? #FHOnline”
They’re asking those “intrusive” questions because they’re gathering data about how people from different origins might have different implicit biases. It’s your call, but since none of the questions are personally identifying, why not help your fellow researchers?
It’s so much easier to give that kind of data when you don’t feel personally threatened by it. As I respond to stuff they could be collecting all sorts of other stuff (IP address alone is a problem… This is America and this is the 21st century… Stuff u do on the internet gets surveiled)
I understand the concern. As a researcher, we have collected zipcode data in order to determine things like demographic and income associated with that area. So it’s fair to question. What’s problematic to me about asking these questions before the exam is that it can bias your response by priming certain social identities before you take the test.