As I mentioned in an early post, I’m taking The Great Courses’ “An Introduction to Psychology” (AITP) hosted by Catherine A. Sanderson, Amherst College.

The second lecture focused on “How and why psychology matters“. It has a decent opening with an overview of main descriptive methods such as surveys, case studies, and naturalistic observations (real-world examples), etc., as well as the limitations such as generalizing from small sample size to larger population, observational influences (Schrodinger!), wrong / inconsistent answers, and/or underreporting (much of the same research limitations for performance measurement and goal-setting). The lecture ended with an overview of indirect observation (mainly using technology), neuroscience, (controlled) experiments, natural experiments, behavioural genetics, and ethical concerns. Overall? The message is it matters because it tells us about ourselves, not earth-shattering.

Interestingly, though, when it was talking about the problems with studying people’s responses, it gave a link to the famous Harvard Implicit Association Test that I realized I had actually never done, even though I’ve seen references to it dozens of times over the years. It’s a great simulation, and I had to try.

The IAT was classically about race relations, at least originally. It starts off asking you to sort words as positive or negative…joy is positive, hate is negative, for example. You press the E key or the I key to sort the words into one of two buckets. Then, it has you sort pictures of white people and black people into two buckets. Totally unrelated, you’re just sorting them the same way, left or right.

But then it tests for a possible / implicit bias — they mix the two samples, and sometimes you see a picture of a person (white or black) or a word (positive or negative). The first time through, you are pressing E for bad OR black people and I for good OR white people. Note that it isn’t a picture of a white person with a good word, it’s a picture of a white person or a black person or a good word or a bad word being sorted into shared buckets. But E is associated with bad OR black and I is associated with good OR white. They do it twice. Then they reverse the combination — now E is associated with bad OR white people and I is associated with good OR black people. The test works by seeing how long it takes you to sort the people and the words into the right bucket (E or I).

In the end, the premise is that if it takes you longer to associate JOY as a positive emotion when the bucket also means BLACK people than when it also means WHITE people, you might have an implicit bias against black people…as it is taking more time for your brain to associate black people and positive emotions when they’re paired.

I confess that I was nervous about taking the test. Would it undermine my own beliefs about my biases? Would it somehow reveal that thinking I was “okay” and not a slave to my physical and emotional genetic contributions from my parents was in fact all a lie? That I actually was a raving racist and didn’t know it? I am, after all, one of those people who has used the phrase “I don’t see colour” without realizing what I was saying. What I really meant, and it’s no better, is that I, of course, see it but it appears no more important to me than cosmetic issues like colour of hair or earrings. I had never realized how potentially disrespectful that was … that it wasn’t relevant to me in how I chose to interact with you, so it didn’t seem relevant to me at all, but was still relevant to you and our interaction.

And yet, I don’t ever remember interacting with colleagues where my reaction was negative and they were all black, for example, but if the bias was implicit, I wouldn’t know. I feel like thinking about it is a Ted Lasso moment…I care that someone is gay or black or transgendered in the sense that it is important to them to be themselves. I don’t “not care”, as Ted says, I care very much. But there is part of too that it isn’t important to me in terms of how I’m going to interact. I hope.

But going to the goal of the test, maybe that muted sensitivity is not simply ignorance or lack of experience; maybe there’s a deeper bias hiding as a result of generally growing up in a white community with very little exposure to black people, and most people in the community were probably at least a little bit racist through ignorance at least. Wow, now I’m having an Avenue Q moment…

After doing the test, it’s interesting that it also collects data on what you THINK the results are before you see the actual results. In my case, I said likely a small/light bias…I’d be hard-pressed to expect anything else based on my background. And that is the exact result it found.

Whil that is somewhat reassuring, I found doing the test disturbing. I found it harder during the second and last part of the test, when I had to switch from E=bad/black to E=bad/white. I made more errors and took longer to classify the answer.

I was also fascinated to see the classic black/white test has expanded its offerings — gender vs. STEM; young vs. old; gay vs. straight; cisgender vs. transgender; abled vs. disabled; gender vs. career choice; Arab/Muslim names; white/black vs. things that look like weapons; fat vs. thin people; religions; skin-tone; and nationalism. Fascinating ideas.

The lecture was relatively basic, nothing revelatory. But the IAT was worth the time investment alone.