Background content: Ideally students would have read some background content on implicit bias coming in to the conversation. One possibility is Kelly and Roedder’s 2008 paper “Racial Cognition and the Ethics of Implicit Bias.”
(Only provide the first half, as it gives some background on implicit bias and empirical evidence supporting its existence.)

Another possibility is just to have the teacher assign them to take an IAT test the night before. Then, one could come in and just discuss some of the main features of implicit bias (outlined below).

Class plan:
1. Show an optical illusion, and poll the class on their responses (which is bigger/longer?). Some options include the Muller-Lyre illusion or the Ebbinghaus illusion.

Inevitably, some students will know these illusions, and answer correctly that they are the same size. Ask some of them (those who answered that they’re the same) to explain their reasoning for answering the way they did. Often, this will include reference to having seen the illusion before, and thus knowing that appearances are deceiving. Be sure to highlight the fact that one still looks bigger, even when they know the trick and are able to adjust their judgments. Let students know you’ll come back to that later.  (5-10 min.)

2. Discuss implicit bias and moral considerations. Review primary features of implicit bias. If no one has done background reading, now is when you might discuss implicit bias generally, including (5-10 min.):

1. that is subconscious (not subject to introspection);
2. that we act on it unconsciously;
3. that it is pervasive;
4. that it affects our actions (lots of evidence that one could cite…e.g., resume studies);
5. that it often contravenes our explicit beliefs (people who espouse egalitarian views often show bias on IAT); and
6. that it concerns many categories (e.g., race, sex/gender, weight, sexuality, etc.)

Ask students to turn to their neighbors and discuss what moral considerations might be at play here. It might help to give examples like: Is someone who demonstrates implicit bias morally bad?; Have I wronged a person if I am biased against her/him?; Do you think someone who has explicit biased attitudes is morally bad? What are the relevant moral differences between explicit bias (e.g. racism) and implicit bias (racial bias)? Can I be responsible for something outside of my control? Encourage students to try to identify their reasons for making particular claims. Reconvene and ask if anyone wants to share. Discuss.  (10-15 min)

Ask each pair to join another pair (groups of 4), and discuss together what kinds of strategies we might use to address implicit bias. Give them some leading questions to consider like: Does media and imagery feed our bias?; If we don’t recognize it, and often are bad at identifying it in ourselves, how can we go about correcting implicit bias? Should we strip all identifying features whenever possible in situations where bias may arise (e.g. resumes)? Reconvene and ask if anyone wants to share. Discuss. (10 min.)

3. Present “deliberate counterbalancing” as a possible solution (this is discussed in the second half of the Kelly and Roedder paper).
For example, say, “Okay, lets suppose I have all of your final papers. I go through and grade all of them according to my best judgment. Then, recognizing that I very likely harbor an implicit bias against X (women, people of color, or some other social category of your choosing), I separate the papers into two piles: Xs and non-Xs. I give all the Xs +3 to compensate for my implicit bias against them.” Ask students what they think of this proposal.

They almost always hate it, citing things like fairness, deservedness, earning grades, etc. Point out that they used this exact strategy (those of them who knew the illusions) in determining the sizes of the circles/lines. You could also refer to viewing something straight (e.g. an oar) when underwater. It looks curved, but we would say that it is straight because we know what submersion in water does to our perception. If we know our perceptions are flawed, how can it be wrong to correct them (he says incredulously)?! This isn’t a moral fix; it’s an epistemological one. Discuss. (This can end up going in a lot of directions, but the students at this point should have thought about it enough to grapple with the issue in a pretty sophisticated way.

The goal is that students have thought hard about implicit bias and the moral dimensions of this issue, and recognized that it is very difficult to determine how to address it or what to think about moral judgments concerning implicit bias. Also,  the hope is that thinking through the issue (and especially the nuances of some of the epistemological and moral concerns at play) will help students identify and scrutinize some of their own underlying assumptions.