Brittle smiles: positive biases toward stigmatized and outgroup targets

We examined individuals' tendencies to exaggerate their positive responses toward stigmatized others (i.e., overcorrect) and explored how overcorrection, because of its fragile nature, could be disrupted. The first 2 studies demonstrate overcorrection: White participants paired with Black partners (Experiment 1A) smiled, laughed, and showed more positive behavior than those paired with same-race partners. Experiment 1B replicated the general effect with a physically stigmatized sample (i.e., facial birthmarks) and then demonstrated that overcorrection is moderated by bias; participants who exhibited more positive behavior toward their partner showed the most physiological "threat" during a stressful task with their partner. We then examined the idea that if overcorrection requires cognitive resources and is effortful, then it may be fragile when resources are taxed. In Experiments 2 and 3, we observed that overcorrection was easily disrupted when resources were compromised (e.g., with stress or cognitive load). Taken together, these studies suggest that positive biases toward stigmatized and outgroup members are fragile and can be undermined when resources are taxed.