A small but impactful shift in the way a group assignment is presented can significantly reduce racial inequality within the group, as well as lead to better work, according to new research by Bianca Manago, assistant professor of sociology at Vanderbilt University. Groups, Inequality and Synergy, co-authored with Jane Sell at Texas A&M University and Carla Goar at Kent State University, appears online in the September 2018 issue of Social Forces.
Previous research has shown that groups often diminish the contributions of minorities, by dismissing their opinions more often, for example, or by being less likely to adopt their ideas. Manago and her colleagues sought to discover whether reframing the parameters of a group task could reduce that inequality, and how that would impact the quality of the group’s work.
“Past research shows that people with different skills working together is good for group performance, but relatively little research has been done on how superficial differences that shouldn’t matter, like race, affect group performance,” Manago said. “We found that when people are more willing to listen to the minority group member, the group does better.”
For their experiment, the researchers put together groups of three volunteers composed of two white women and one Mexican-American woman to work on a problem-solving task once a week for three weeks. Each week, they were asked to rank the importance of 12-15 of scenario-specific items to survive in one of three dangerous places: the desert, the sea, and the moon.
For each week’s scenario, each volunteer was instructed to first create her own ranking, then work with the other two women in her group to develop the ranking they would submit as their collective answer.
On the first day of the experiment, the control groups were told that some participants would be better at the task than others, and that the researchers were studying what makes some groups more successful than others.
The experimental groups were told something slightly different: They were told that the task required drawing from a wide variety of skills and that no single person in a group was likely to possess all of the abilities required to succeed. This, the researchers theorized, would shift the volunteers’ expectations about their own competence as well as their groupmates’.
Each week, the researchers not only evaluated the quality of the group answer, but compared it to the individual answers to see how well individuals performed against the group, as well as see who changed their minds. This allowed the researchers to measure deference—the frequency with which an individual changed their minds during a disagreement—as well as synergy—the ability of a group to outperform any single individual in the group.
In the control group, the white participants consistently exhibited the lowest levels of deference—they were much less likely to change their minds to agree with the Mexican-American group member than the reverse. In the experimental group, however, where the participants were told everyone had something of value to contribute, the white women deferred more frequently to the Mexican-American women than they did in the control group.
Interestingly, the researchers noted, this did not hold true for one of the tasks: the lost-at-sea scenario. Feedback from the volunteers suggested that this was an especially difficult task—likely because a number of the seafaring items in the list were unfamiliar, such as a sextant, and therefore difficult to rank. “In that case, we believe the presence of uncertainty, unfortunately, caused people to fall back into old habits,” Manago said.
Finally, the researchers found that the experimental groups outperformed the control groups. By the end of the experiment, the experimental group was 40 percent more likely than the control group to achieve some synergy, and 20 percent more likely to achieve a lot.
“That was a really cool finding,” Manago said, “because not only does it say that diversity is good for diversity’s sake, it says diversity improves us and makes us better as a team.”
The research was supported by NSF Grant 0961940.