:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/40142794/97783394.0.0.jpg)
As America moves toward accepting more video evidence from surveillance, dashboard, and body cameras in courtrooms, researchers are finding that these relatively new technologies come with some serious drawbacks and caveats.
Studies have already shown that video evidence isn't always enough to cut through jurors' biases — toward either police or people of certain races and groups — when judging a case. Now, a new series of studies from New York University and Yale researchers suggests that ambiguous video evidence may actually reinforce people's biases if looked at in a certain way.
"We wanted to really understand how people can look at the same evidence and come to completely different conclusions," says Yael Granot, a researcher at NYU and lead author of the study.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/1403162/Justice_Is_Not_Blind_study.0.png)
The chart shows that people who have a "weak identification" with police officers prefer harsher punishments than people who have a "strong identification" when they focus more on what a police officer is doing in video footage. ("Justice Is Not Blind: Visual Attention Exaggerates Effects of Group Identification on Legal Punishment")
In two studies, the researchers used a questionnaire to gauge 152 participants' biases toward police, then had the participants watch two videos of altercations between a police officer and a civilian in which, based on independent reviews, police wrongdoing was ambiguous. Participants, whose eye movements were tracked with sensors, weren't told where to look for the first video, but, for the second video, some were asked to look at the civilian and others were told to focus on the officer.
Researchers found participants with a bias against police saw the officer's actions as more incriminating when they watched the officer more closely. Instead of developing a more objective view of the officer's actions by playing closer attention to his actions, participants doubled down on their initial ideas.
"The more people looked at this other group member in this ambiguous sight, the more their identification started to polarize judgments," Granot says. "People who were anti that group were much more likely to punish than people who were pro that group or strongly identified with that group."
To make sure the findings didn't apply only to police, researchers conducted a third study in which participants watched a fight between two college-aged white men: one wearing a blue shirt, and the other wearing a green shirt. Before the viewing, participants took personality tests and were told which group — blue or green — they most likely belonged to. Again, those who focused more on the person who didn't belong in their designated group tended to find his actions more incriminating.
Of course, these findings only apply to ambiguous video evidence. It's likely, Granot cautions, that straightforward footage wouldn't let people's biases come into as much play. "Your expectations and your biases can only fill in the blanks for you when there are blanks," she says.
Granot says the study's findings, along with further research, could help build better instructions for jurors and judges looking at video evidence — perhaps by encouraging them to look at video footage more holistically and take a broader view of the scenes.
"This isn't about the video evidence itself," Granot says. "It's how we watch it."
