For example, in a recent study, participants were asked to read one of four news articles about a hypothetical African American man. The first story was about a professor receiving an award. The second was about an athlete, a role stereotypically linked to African Americans. The third was about a white-collar criminal and the fourth about a violent criminal. No matter which story they read, all participants saw the same photo of the man.
Despite that fact, when participants were asked to reconstruct the man's face on a computer, they gave the man more stereotypically African features, such as darker skin tones, if he had been portrayed negatively as the white collar criminal, or--especially--as the violent criminal.
The fact that they didn't associate more African features with the athlete, who was portrayed positively, suggests the participants linked these features to negativity rather than stereotypes. And this happened regardless of the participants' self-reported racial attitudes, which means that they weren't even aware they were having these troubling responses.
If so, then it's not enough to study the way the media portray racial groups. It's also important to study how the public interprets and remembers--or misremembers--the news in a way that sustains racism.