KU studies: Media literacy can reduce stereotypes; mass communication research samples lack diversity
LAWRENCE — The media that people consume inform a large part of their everyday life, including how they view others. A pair of recent studies from the University of Kansas shows that a media literacy intervention can help reduce stereotypes people hold about Black Americans and that a majority of journalism and mass communications studies have lacked diversity with overly homogenous samples.
Media literacy and racial stereotypes
Joseph Erba, associate professor of strategic communication at KU, conducted a study examining the effects of a media literacy intervention on racial stereotypes. He found that when readers read a news story about sports, drugs or welfare — topics that disproportionately stereotype the Black community — those who took part in a media literacy intervention held fewer stereotypes than readers who did not. Erba and Yvonnes Chen, associate professor of strategic communication at KU, had previously conducted research on a media literacy intervention with students.
“We saw that those interventions were successful in lowering stereotypes in students, especially about African Americans and Latina/os,” Erba said. “But they were face-to-face, one group at a time. We wanted to look at an intervention that could be done online with a larger sample.”
Erba conducted an online experiment with more than 700 white Americans who read a news story on a fictional news website about sports, drugs or welfare. Participants were given one of the three stories at random, and half were given further context to read — the online media literacy intervention — and half were not. All then answered questions about their attitudes toward Black subjects in explicit and implicit terms by agreeing or disagreeing with statements from the Pro Black Attitude Scale and the Symbolic Racism Scale, two previously validated research instruments.
The findings showed those who received the intervention held fewer stereotypes than their counterparts. Those who read the sports story showed the greatest difference in stereotypes, followed by the story on drugs, while those who read stories on welfare showed no difference. The intervention included further context for readers in all three areas. One example for the sports story included information on educational inequalities in the United States and how many Black Americans had to hope for an athletic scholarship to be able to continue their education after high school. The crime intervention included information about how members of the Black community are more likely to be arrested for drug offenses, even though they do not consume or sell drugs at higher rates than white Americans, and the welfare story included information about the difference in welfare benefits and income inequality between Black and white Americans.
The unequal reductions in stereotypes between topics could be due to the level of stereotype associated with each in audiences’ minds, but the fact that the intervention did show reductions is encouraging, said Erba, who presented his findings to the International Communication Association in 2021. He also noted that those who received the intervention were asked their opinions on reading the news with the additional context included, and they reported enjoying reading the news more than those who were not provided the context.
“Journalists tend to see themselves as objective observers and don’t want to appear biased. So we wanted to see if having this additional context affected how people viewed the news source,” Erba said. “Not only did the intervention work, but there was no difference in the level of trust, and participants enjoyed the news even more when they got it. That’s a win-win as far as I’m concerned.”
Erba continues to test the online media literacy intervention with college students to see if a different sample produces different results. In a time of racial tensions, unrest and increased calls for social justice, a media literacy intervention that can reduce stereotypes holds promise, especially when research has shown the Black community is overrepresented in news coverage as perpetrators of crime, underrepresented as victims of crime and in other negative lights, he said.
“What we have very little information about is how we can intervene to mitigate those effects,” Erba said. “Media literacy is a very rich field. We have research on media’s effects on behavioral measures like smoking and drinking, but very little on race.”
Who are the masses in mass media research?
Another KU study found that mass media research of the last decade-plus has depended on samples lacking diversity, which can call into question how much researchers truly know about mass media’s effects on the population.
Erba; Peter Bobkowski, associate professor of journalism & mass communications at KU; Brock Ternes of the University of North Carolina-Wilmington and former KU graduate student; and Yuchen Liu and Tara Logan, former graduate students at KU, wrote a study in which they conducted a census of 1,278 mass communication studies, all of those published in six high-impact communications journals from 2000 to 2014. They found that when the studies did report on the demographics of their samples, they relied heavily on young, educated white women. The study was published in the Howard Journal of Communications.
In the 2000 U.S. Census, Latina/os were the largest minority for the first time. The U.S. population has diversified even more since then, yet many studies that form the basis of knowledge on media effects in America did not have samples that mirrored that diversity.
“We’re taking a step back and saying, ‘We have these results, but how did we get there?’ This study shows we may not know as much as we thought we know about media effects, because we’ve based them on a series of studies in which the sample may be too limited,” Erba said.
The American Psychological Association recommends that researchers record and include demographic information about their study samples including age, gender, race, income and education levels. Several of the studies included no information on demographics, and authors point out that demographic reporting did improve over the 15-year study period. Three-fourths of the studies reported gender, two-thirds reported age, two-fifths reported race/ethnicity, and one-third reported education. While the reporting improved, the authors point out that when the information was included, it showed the majority of study participants were young, with an average age of 30; 60% were women; the vast majority were white; and most had at least 13 years of education. Therefore, the average participant was not representative of the larger U.S. population.
Years of research have pointed out the importance of how a person’s identity, including age, gender, education, race, income and other factors, influence how they interact with mass media in areas such as health communications and news coverage. However, when samples are not diverse or too homogenous, it limits the implications of the studies for the broader population.
“I hope that this study adds to the conversation happening across the social sciences about how inclusive our disciplines are in their faculty ranks, research and publications,” Bobkowski said. “This study contributes one more data point in support of critically evaluating our collective work and committing to do better.”
The authors argue that not only does mass media scholarship perhaps not know as much as previously thought about media influence but that researchers should actively take steps to reverse the trend. First, they recommend recording study participants’ demographics and exploring potential differences between participants representing different demographic intersections. They also challenge researchers to be more conscious about sampling techniques by recruiting more diverse samples and remembering that people identify with social identities on different levels.
“One lesson from this study is the need to improve how mass communication approaches social identity. Measuring participants’ identities and identity salience adds only a few minutes to a study, and reporting these variables adds only a few sentences to a manuscript,” the authors wrote. “Exploring the role these variables may play vis-à-vis the media makes a study’s results more transparent for readers and may lead to new insights. Mass communication researchers should reflect on their participants’ identities and critically assess what their results reveal and, most importantly, do not reveal.”