Journal article critique; media skepticism & trust

Tsfati, Yariv & Cappella, Joseph N. (2003). Do People Watch What They Do Not Trust? Exploring the Association Between News Media Skepticism and Exposure. Communication Research, 30(5), 504-529.
http://crx.sagepub.com.ezproxy.lib.utexas.edu/cgi/reprint/30/5/504

This paper explored the relationship between media consumption and media trust – specifically to investigate whether media trust affected consumption patterns. The hypothesis suggested that people who trust mainstream media consume it; those who don’t seek alternatives.

Assumptions:
The author uses an elegant, parsimonious definition of media skepticism; “a subjective feeling of mistrust toward the mainstream news media.” Specifically, this involves questioning credibility and reliability, whether the MSM is right and fair. But this is a two-part study, with the other key aspect being a hostile media effect. This paper assumes rational audiences, a la the rational self interests of consumers as described in the Media Economics book. This is the first of four assumptions made by the author, each of which lacks a bedrock foundation. There is also the assumption that consumers desire “accurate knowledge about the nonimmediate world;” that they “have an incentive to ignore many stimuli; and that, for convenience of analysis, there are two types of media, mainstream and non-mainstream. With deference to the elegance of analysis of these assumptions, I believe each of these lacks fail to capture the complexity and diversity of contemporary media and those who consume it. For example, the study equates PBS and MSNBC as mainstream; while I would agree, I dare say dissenters could be found. Likewise, online news and political talk radio are equated as non-mainstream sources.

Two of the hypotheses seem fairly obvious:
H1: Mainstream media skepticism will be associated with lower mainstream news exposure.
H2: Mainstream media skepticism will be associated with higher non-mainstream news exposure.

The final hypothesis seems less explored and more interesting.
H3: Skepticism will be associated with news media diets: the higher the skepticism, the higher the non-mainstream component in audiences’ media diets.

I also take issue with the data set. The Electronic Dialogue project offered free WebTV units to those who volunteered to participated in surveys and online political discussions. As an early attempt to gather public opinion data online, the ED study is laudable. But the panel inherently lacked early adopters of online technology and likely excluded both those who were pioneers in consuming information online as well as online laggards. Cudos to the author for attempting to validate the data with cross comparisons to two well-respected and established data sets; the National Election Studies and the Annenberg Public Policy Center.

Analysis:
The analysis was well done. Nine factors were scaled to create a media trust index and factor analysis clearly delineated between mainstream media and non-mainstream, which included talk radio and the Internet. While skepticism has already been voiced about lumping Internet with alternative, the factor analysis found only 15% correlation between the two, validating the data over my objections. The author found that a majority of audience members still preferred mainstream media and that most subjects were skeptical of mainstream media – especially national television news and newspapers. This seems to contradict both hypotheses 1 and 2, which suggested those who watch mainstream news more will be more trusting of it; and those who don’t trust it will avoid it. In fact, given the author’s definition of mainstream versus non-mainstream, political ideology was as much stronger predictor of media skepticism than was news consumption habits – although skepticism also was associated with listeners to talk radio. Therefore H3 was supported and, in fact, political interest was the greatest predictor of any and all media consumption and the skeptics seemed to cast a wider news net than non-skeptics.

Of course this study was conducted in 2003 and, hence, suffers from the inability to consider the wealth of academic research conducted in the ensuing years – but, pretending this was a contemporary study, I would have suggested the new fall, 2009 Pew study examining media credibility and trust which found, as I described, that political ideology was a large predictor of trust in the mainstream media. Also, studies last year and this year by Gil de Zuniga, Valenzuela, Kaufhold, Bachmann, Shah, Eveland and many others have constructed new ways to define traditional, online and alternative media, including recoding online MSM sources with traditional; blogs as alternative; and treating conservative talk radio as an outlier variable – highly predictive but incomparable to other media types.
I would certainly recommend that this article be revised and resubmitted, with more careful thought given to how the literature leads to the hypotheses and to more contemporary literature and data sets. This study was well thought out. It could benefit from a more thoughtful embrace of the definitions of mainstream and non, traditional and alternative.