The operation of source-credibility effects — roughly, that I may believe something if X says it but discount it if it comes instead from Y — is well established in social psychological research on attitude formation and change. But here’s an interesting new twist, in the context of current controversies over “fair and balanced” political coverage by the media.
In a study reported in the latest issue of Political Behavior (abstract here), Joel Turner notes that many Americans perceive TV news coverage as ideologically biased. As cases in point, Turner singles out the widespread sense that the Fox News Channel tilts to the right while CNN leans left.
Studies of bias in reporting of the news typically “content analyze” news reports from various media outlets to determine whether they are presenting more or less the same picture of current events. The bias hypothesis is sustained to the extent that such content is found to vary across outlets.
Turner took a different tack. He deliberately held content constant and looked for another type of bias — bias among TV viewers based not on what they were seeing and hearing, but on the channel they (thought they) were watching.
To check out this possibility, Turner conducted a simple experiment. He selected five stories from Fox and CNN and told his experimental subjects (predictably, college undergraduates) that they had aired on either Fox or CNN; he also played these stories for a control group that was told nothing about their source. The subjects in each group then scored the ads in terms of their ideological bias.
The results? The stories that Turner attributed to CNN were perceived as significantly more liberal than those that received no network attribution. By contrast, when Turner attributed the very same stories to Fox News, they were seen as significantly more conservative than their unattributed counterparts. These source-based perceptual biases were especially pronounced among viewers who didn’t share the putative ideological perspective of the network they thought they were watching. That is, conservatives watching CNN were more likely to perceive liberal bias than were moderates or liberals, and liberals watching Fox were more likely to perceive conservative bias than were moderates or conservatives — again, in the very same stories.
None of this means that allegations of biased coverage are incorrect, of course. It does, however, suggest that such bias exists, to a significant degree, in the eye of the viewer rather than, or in addition to, in the contents of the news stories themselves.