So there’s some evidence that stress doesn’t help us make good decisions. But there may be exceptions, particularly when it comes to evaluating potential threats. In a new study (which you can read in its entirety here), researchers subjected police officers to a stressful situation, raising their cortisol levels. They then had them play a video game during which the officers had to decide whether to shoot the men who appeared on the screen. Some of the men were black, some were white. Some were armed, some were not.
The researchers found that higher stress levels helped the officers (no matter what race) make better decisions about whether to shoot, but only when the fictional assailants on the screen were black. That seemed odd, so I asked one of the authors, Modupe Akinola, to explain.
Q. I would have assumed that an increase in cortisol levels brought on by stress would lead to poor decision making (and, as you mention, there are studies that indicate that stress can cause people to make riskier decisions). But in this case, officers under stress made fewer errors. Any idea why this might be?
A. It is important to think about the nature of the decision-making task. In our study, we used a threat-related decision-making task. Some studies have shown that when individuals are highly responsive to stress (i.e., they show large cortisol increases in response to stress), then they are attentive to threat cues and are vigilant to danger. These studies suggest that in the context of threat-related decision making, stress might enhance task performance, which is what we saw in our study. However, in other types of decision making that may not require vigilance or attentiveness to threat, we might expect to see that stress can have a different effect.
Q. You found that “the relationship between increased cortisol reactivity and fewer error rates in the decision-making task was stronger when the targets were armed and black than when the targets were armed and white.” So officers made fewer errors when under greater stress, but only when dealing with black men who had guns. Were you surprised by this? What do you make of it?
A. There have been some extremely vivid accounts in the media, such as the shooting of Amadou Diallo in the 80s, in which unarmed black men were shot by police officers as they were mistaken for carrying guns. These real-life situations suggest that officers may make errors when dealing with black men relative to white men, particularly when they are unarmed. So one surprising finding from our study is that under stress, officers were no more likely to mistakenly shoot a black target than a white target.
The fact that officers made fewer errors when shooting armed black targets is actually consistent with what we might expect to see. Since there is a widely shared societal stereotype linking blacks to danger, just seeing a black target can trigger an automatic response to consider this perceptual cue as dangerous. Moreover, in the case where the target is armed (or paired with a danger-related perceptual cue), the correct response to shoot is facilitated by the stereotypic association. So the fact that we saw fewer error rates when the targets were armed and black was consistent with our expectations.
Q. So the stereotype is actually helpful?
A. While it might appear that the stereotype is helpful when examining error rates for armed black targets, when you look at error rates for armed white targets, it becomes clear that the stereotype can be rather harmful. Officers did not correctly shoot armed white targets. White targets were not presumed to be potentially dangerous, so as we see in our results, officers made more errors when white targets were armed, which is certainly harmful as it could put officers’ and civilians’ lives in danger.
Q. Based on these results, is there anything officers (and others) should do?
A. One of the key questions that these results raise is: What can police officers and police departments do to increase accuracy in shooting decisions? Given that stress-induced cortisol increases were generally found to enhance accuracy, it is possible that modulating stress levels in training could increase overall accuracy in error rates for both black and white targets. Future research should explore the degree to which repeated exposure to Firearms Training Systems that provide naturalistic simulation environments, and enhancing the levels of stress experienced in these simulations, can help to increase overall accuracy in shooter simulations like the one used in this study. These and other implications have been discussed with the police department in this study.
