Monday, November 14, 2011

Brain scans measure racial bias

Brain scans measure racial bias

Researchers at Yale Law School have discovered that brain scans may better predict jurors' racial bias than previously established methods of testing.

In a new study published in the journal Social Neuroscience last week, Yale researchers studied the correlation between compensation in employment discrimination cases and brain activity during tests for racial bias. The study concluded that functional Magnetic Resonance Imaging (fMRI) scans may better predict verdict size than the traditional tests. The results may be significant to the developing field of neurolaw in providing a way to ensure an unbiased jury — a Constitutional requirement — but outside psychologists said the study's significance is limited by its small sample size.

"This is a novel result, obtained with novel methods," Tony Greenwald '59, professor of psychology at University of Washington, wrote in an email to the News. He added that the researchers established a new method to analyze the fMRI results which looked at the entire brain instead of just specific "regions of interest."

Researchers asked study participants to match images of black and white faces with positive, negative, and neutral adjectives, according to varying sets of rules. For example, they might be told to match all "black faces and positive adjectives" as quickly as possible.

In the Implicit Association test (IAT), which has been used since the 1990s to reveal subconscious biases, researchers measure the reaction speed with which participants match the faces and adjectives such as "good" or "bad", with a longer reaction time signifying a bias against associating the two terms. For example, a subject biased against blacks might take longer to match a black face with a positive adjective.

The fMRI measured the neural activity of participants while taking the test, tracking changes in blood flow across brain regions. Researchers then correlated these results with the money subjects awarded victims of employment discrimination in theoretical cases.

The study found that the variation in fMRI brain imaging results had a higher correlation with the verdict sizes than the IAT test result variation did.

"Our study demonstrates that fMRI measures might have more predictive value than commonly used behavioral measures such as the implicit association test." Marvin Chun, professor of psychology and co-author of the study, said in an email to the News.

The 1964 Civil Rights Act prohibits employment discrimination on racial grounds and provides discrimination victims a judicial means of redress through the Equal Opportunity Employment Commission, which can file lawsuits on behalf of employees. The monetary compensation is often determined by a jury, and the verdict size may be influenced by the conscious and subconscious biases of the jurors, said Harrison Korn '11 LAW '14, a co-author of the study. Korn is a former associate managing editor for presentation at the News.

"We are not suggesting that people go out and start scanning jurors, but it does raise the issue that unconscious bias is a problem and we should be looking for ways to counteract it," Korn said.

He added that the high cost of fMRIs — almost $1,000 per participant in this study ­— and the perception of neural scans as invasive make it impractical to scan each potential juror in the jury selection process.

The study emphasizes, however, that the American legal system must provide unbiased juries in order to ensure due process of law. In an email to the News, Director of the MacArthur Foundation Research Network on Law and Neuroscience, Owen Jones LAW '91 said that the while the results of the study will not revolutionize juror selection, they may help researchers to develop better understandings of human bias. For example, in a 2007 New York Times article, Jones suggested that lawyers could use brain scans of potential jurors to exclude those who were unlikely to be sympathetic to their argument.

"Anything that helps us to understand the mechanisms of racial bias might help us to develop better systems for identifying it, combatting it, and minimizing its effects," Jones wrote in the email.

Brian Nosek GRD '02, associate professor of psychology at University of Virginia, said he questioned the "predictive validity" of fMRI studies due to their limited sample sizes, which are often a result of the high cost of brain imaging. After eliminating six subjects because of technical difficulties, the Yale study was comprised of 19 white, non-Hispanic subjects between 18 and 26 years of age.

Other scientists were similarly wary of extrapolating from the results before they were confirmed by larger studies.

"It's a result that needs corroboration by further research before one should venture any confident interpretations." Greenwald wrote.

In fiscal year 2010, the United States Equal Employment Opportunity Commission received almost 100,000 reports of employment discrimination.

No comments: