UNL uses video games to study racial bias
By Jacob Bryant, NewsNetNebraska
The University of Nebraska-Lincoln’s cognitive psychology department has just finished analyzing data on the effects video games can have on racism and sexism. The university is using video games to detect people’s sexual and racial biases.
“They would play the Kinect with someone that looked like them or someone that didn’t either race or gender or opposite race or gender,” UNL grad student Leslie McCuller said. “We would collect more scale information, and then the IAT, which looks at how strongly you link social information to concepts of race or gender.”
According to professor Mike Dodd, IAT stands for Implicit Association Test. The IAT is a test of connections a person has.
“On the screen you would see faces of white people, faces of black people, and positive and negative words,” Dodd said, further explaining IAT. “We tell you hit a left button if it is a white face and good word, or the right button if it is a black face and bad word. Then we do left for white and bad and right for black and good.
“What you tend to find is that people are fastest to make responses when you pair white with good and when you pair black with bad. It is something that is conditioned over time, but doesn’t necessarily mean you have a bias or a stereotype.”
Using video games as alter egos
To do that with the video game, McCuller and Dodd used the Kinect game Kinect Adventures. The game allows the user to create an avatar personality to use when playing the game. Test subjects could play as avatars that resembled their race and gender, and then use avatars that didn’t resemble them.
When it came to their study’s conclusions, McCuller and Dodd said that test subject reaction times were much different.
“If they saw the white/good pairing against the black/bad it took less time for them to say the black/bad pairing when they played with the in-group avatars as opposed to the out-group.”
McCuller said that was good because in most cases it is the other way around. It also dropped the amount of errors that were being made
as the subjects rotated from in-group avatars (same race or gender) and the out-group avatars (different race or gender).
“These are just small effects but playing with the out-group avatar for just a brief period of time changed the way they responded bias wise,” Dodd said. “The notion was that, through this experience of almost walking in another person’s shoes of the out-group avatar, it lowered the base line for the more implicitly bias response.”
“Essentially, you weren’t as quick to say black was bad and white was good,” McCuller said. “You also didn’t make as many errors, because with the white/good black/bad thing people often mess up on that. They assume black is bad and click on the wrong label. They made fewer errors playing as an out group avatar.”
The small effects McCuller and Dodd recorded during the experiment were seen after only 20 minutes of playing the game. Dodd and McCuller believe that playing the game for longer would lead to more pronounced effects. The decrease in errors was evident in the facial errors. Subjects made fewer errors when the good/bad questions related to faces but not for questions relating to good and bad words.
McCuller is also analyzing data for a third group that she wanted to cover; age.
“We were trying to hit three potential sources of prejudice,” Dodd said. “Depending on how the results turned out, we could write a publication on this.”
McCuller is using the research as her PhD data. If the remaining data shows as much promise, than the two plan to look for publication options of their findings.
“We didn’t expect anything,” Dodd said. “This is based on a notion that people like other people when they replicate your actions. If you ever find yourself in a conversation with someone who is speaking slowly and you’re speaking slowly back to them even though it isn’t how you normally speak, it’s just a tendency that we have. We tend to replicate other people because it makes them like you more.”
Dodd thinks the research he and McCuller did was a huge jump forward for what video games studies have done at any other point.
“Video game studies used to focus on how bad video games are,” Dodd said. “The basic cognitive stuff has been on low level attention. This is the first study I know of to look at social behavior.”