If words don't get the message across, actions just might do the job instead. After establishing that subtle nonverbal pro−white biases are prevalent in popular television shows, Tufts University Interpersonal Perception and Communication Laboratory researchers recently found that these nonverbal behaviors directly influence racial biases among television viewers.
"Subtle race bias exists even as more explicit bias diminishes — that's been repeatedly demonstrated in laboratory studies," Research Assistant Professor of Psychology Max Weisbuch, the lead author of the study, said. "We wanted to look at the subtle social influences that can affect people outside of the lab, and television is a domain that reaches a lot of people. It was a matter of taking what some lab studies suggest ... and investigating it in the media."
Co−author Kristin Pauker, a visiting researcher of psychology from Stanford University, credited Weisbuch with the conceptualization of the study, entitled "The Subtle Transmission of Race Bias via Televised Nonverbal Behavior." Science Magazine published the study in December of last year.
"I thought the idea for the study was really interesting because we know that there are racial biases that people still hold," Pauker said. "So it's been established that racial bias exists, but what our study showed was that racial bias can actually be transmitted via popular TV shows through something as subtle as nonverbal behavior."
According to Weisbuch, the research began in the winter of 2007 and took place in the Interpersonal Perception and Communication Laboratory. Directed by Professor of Psychology Nalini Ambady, the research was comprised of four interrelated studies, each building on the next. The studies examined nonverbal biases, a subset of nonverbal behavior defined by Weisbuch as "the tendency for people to express more positive nonverbal behavior to one group over another group."
Two experimental conditions were created, examining nonverbal bias behavior that was classified as pro−white or pro−black. Silent clips were chosen from 11 popular scripted television shows that feature prominent white and black characters, including "Friday Night Lights," "Grey's Anatomy," "House" and "Bones."
"It was hard to find TV shows that fit the criteria that had major characters of different races," Pauker said. "We needed black and white characters who were fixtures throughout a season, which was a problem — when characters on a show die off, they tend to be the black characters."
To control for possible confounding variables, the researchers matched the black and white characters on several characteristics, such as status, intelligence and attractiveness. "We wanted to assure, if we found an effect, that the effect was due to race and not to some other variable," Weisbuch said.
Once the shows were chosen, the researchers randomly selected three episodes of each show. They then carefully examined those episodes to find clips of the characters that contained at least 10 seconds of interpersonal interaction, among other predefined criteria.
The researchers then removed sound from the clips and chose the segments that demonstrated the most extreme pro−white and pro−black nonverbal bias; in other words, the clips in which the characters were on the receiving end of extreme positive or negative nonverbal behavior. The final pro−white and pro−black conditions each contained 50 five−second silent clips. The study participants were screened ahead of time to ensure that they had not seen the television shows in question.
Weisbuch emphasized that although the television shows did exhibit much stronger and more frequent instances of pro−white bias, the shows are some of the more diverse programs available, as demonstrated by their commitment to the inclusion of racially diverse characters.
"The shows we used to find this pro−white bias are actually the shows that really do put in effort to be diverse and to represent everyone," Weisbuch said.
Pauker highlighted the fact that the television shows were not cherry−picked to support the researchers' hypothesis.
"It's not like we went through and picked TV shows that we thought would display the most bias," she said. "We picked popular TV that fit our criteria and that lots of people watch."
The researchers predicted that the participants who viewed the pro−white clips would themselves exhibit a stronger pro−white bias as measured by the race−based Implicit Association Test, which measures unconscious associations.
The researchers found that brief exposure to a television character's nonverbal bias did influence the participant's own biases. "We found what we expected," Pauker said. "But when it came down to it, when we had the results in front of us, we were surprised to find such confirmation."
Pauker also noted that the participants were unaware of the bias patterns they had unwittingly internalized. According to Pauker, the participants responded at chance levels to a questionnaire examining recognition of bias patterns across the pro−white and pro−black clips.
"That's a huge problem, because if your guard is down and you don't realize that this bias exists, then you lack the ability to self−correct," she said.
Weisbuch agreed that subtle racism is problematic because it can affect people without their awareness. "There's the definite possibility that popular TV might currently be playing a role — a fairly sneaky and insidious role — in perpetuating bias," he said. "At this point, by publishing, we're hoping to raise awareness."
However, Weisbuch cautioned against sensationalizing the results. "It appears that many popular TV shows exhibit race bias and that exposure to this bias influences viewers," he said. "However, we do not yet know the extent to which such exposure exerts a culture−wide influence."
Both researchers hope to follow up on their findings with future research. Pauker is considering performing similar studies with children, while Weisbuch plans to examine the cognitive and communicative processes involved in nonverbal bias transmission.
"I'd like to try to isolate the processes — the mediators and moderators — that account for the effect," Weisbuch said. "That would help us to get an idea of factors that might inoculate people against the influence."
