This presentation was given Saturday at the AEJMC Annual Conference as part of a refereed paper research session on “Bias and Branding” sponsored by the Radio-Television Journalism Division. This is a rudimentary script to an oral presentation with visual aids, so it will leave a bit to be desired for the blog reader.
For the PowerPoint accompanying this presentation, email Dylan.
For more on the conference as a whole, view my main post on AEJMC ’11.
Claims of media bias are nothing new. And despite plentiful content analyses that show little to no evidence of some collective attempt to mislead the public, perceptions of bias not only remain, but have increased dramatically over the past decade, a time frame that correlates with the rise of Fox News and MSNBC – cable news networks to compete with CNN and create a competitive marketplace.
Researchers like Sutter or Anand, DiTella, and Galetovic have looked at news coverage economically, and cable news as just what it is – a for-profit industry. That means, like any differentiated product market, the news outlets must seek a place along a continuum of potential audiences. With radio stations, it would be genres of music; with news it could be the types of stories covered (intl/domestic; hard news/entertainment), but we often think of it in terms of political ideology.
The suggestion is that our oft-idealized paradise of objectivity doesn’t make good business sense, because wide-open market segments are left untapped while everyone battles for the middle. If every station in town is playing country music, why don’t you try reaching out to the hip-hop fans? (Ideology isn’t quite so drastic, but you get the idea.)
So instead, content analyses (or a casual channel surf during primetime) have suggested that the cable news environment looks something like this. Each network has differentiated, targeting its own particular audience. Which begs the question… How do these differentiation attempts influence the audience’s perceptions of bias in those networks?
Two research questions were asked. The first is in the traditional vein of hostile media phenomenon research. The supposition is that stronger partisans perceive greater bias against their position than weaker partisans. The second RQ deals directly with these branding techniques and cueing. If MSNBC, for instance, has moved itself to the left side of the continuum, does the content of the message matter, or do audiences make judgments based on the source alone?
An experiment was conducted with a convenience sample of 324 student participants from a community college and university in a mid-size southeastern metropolitan area. Each participant was given the same article, specifically designed to be unbiased. Each article displayed a different news logo: MSNBC, CNN, Fox News, or no logo, to serve as a control. A questionnaire was attached asking participants about their perceptions of the article, the networks themselves, and their personal ideology.
This is what the article looked like.
The topic was a climate change bill that was actually being considered in Congress. Large portions of the article were factual, though some details were fabricated to suit the experiment’s goals for balance and avoidance of distractions (like a quote from a well-known divisive source). There were an equal number of pro/con paragraphs, quotes, and factual citations.
The topic was of importance. It needed to describe a clash between parties, in order to measure the degree to which political ideology affected perceived bias. However, the topic could not be one of intense ideological divide and dialogue (like immigration or abortion rights) due to the risk of the reader’s personal beliefs about the topic overwhelmingly outweighing perceptions regarding the content of the article or the news network to which it was attributed. Climate change seemed to be a topic that struck the correct balance for the purposes of this study.
So many studies on hostile media phenomenon and perceived media bias use topics outside the political realm (agriculture, sports, travel). But virtually all claims of bias in the media relate to political news. I consider the use of a timely political story was one of the distinguishing elements of this research.
What happened was in stark contrast to previous research.
Participants perceived the article to be unbiased, regardless of their personal ideology or the presence of a network logo. It is important to note that a large number of participants (about a third) reported being “not sure” of the article’s position. Those responses were removed for this dataset.
Breaking it down by ideology of the participant, conservatives perceived the least amount of bias in their favor (which would suggest a relative hostile media phenomenon), but liberals perceived the most in their favor – completely contradictory to the theory. Additionally, the mean responses were remarkably similar.
So it would appear that the network logos did not have an impact. But, did the participants have opinions about the cable channels in the first place? If they didn’t, no wonder the logos didn’t matter.
But the data suggests that they did have perceptions related to the networks. Conservatives reported a large amount of liberal bias in CNN and MSNBC, and didn’t watch the networks all that often. Liberals saw them as more balanced, and watched them more.
Interestingly, liberals, conservatives, and moderates alike all reported a conservative slant in Fox News coverage, yet it was the most-watched, in line with the ratings. (The difference between liberal and conservative viewing of Fox News was just under the threshold for statistical significance).
Let’s make this easier to see. There was a clear, stacked difference in the perception of CNN and MSNBC. This figure displays the relationship between perceived ideology of a network, and frequency of viewing. Those who perceived the most bias in CNN and MSNBC were also the ones who watched them the least, which suggests that audiences are recognizing this differentiation, avoiding disagreeable sources, and labeling them as biased based upon secondhand sources.
All of this suggests that the participants had distinguishing perceptions of the differentiated cable news networks. Yet, those perceptions did not carry over to the experimental article. Why?
A text article is not the same as a television broadcast. The focus required to read the article may have overpowered the ability of the logos to serve as effective cues, though similar designs have been used before with more typical results. And, of course, college students are not the most politically active, nor news-consuming demographic.
Still, two generalizations: This study saw audiences that flocked to agreeable sources while avoiding disagreeable sources and labeling them as biased without actually consuming them. A dangerous selective exposure that supports assimilation biases and enables a partisan to insulate oneself from reality.
This study also saw the same participants that reported such behavior to disregard partisanship and perceptions of the cable networks and perceive an article designed to be unbiased as just that. A great hope for the next generation of news consumers in an increasingly differentiated news environment.