Has Newspaper Credibility Mattered? A Perspective on Media Credibility Debate

Article excerpt

At least three studies have suggested that newspaper credibility matters partly because low assessments of newspaper credibility are associated with low levels of newspaper use and of support for the press' First Amendment rights. But a reanalysis of the studies' data found no relationship between newspaper credibility and frequency of newspaper reading. Furthermore, doubts have been raised about the linkage between credibility and support for free expression.

A limitation common to these studies may help explain their apparently conflicting results. All were based on cross-sectional survey data and therefore could not detect patterns over time. Newspaper credibility's relationship with newspaper reading and support for free speech rights may have ebbed and flowed over the years, appearing in some studies but not in others. This study brought a historical perspective to the debate by inspecting two decades' worth of General Social Survey data for evidence of relationships between newspaper credibility on one hand and newspaper reading and support for free expression on the other. Like many secondary analyses, though, this one had to rely on measures that did not correspond exactly with the concepts being investigated.

GSS surveys have not consistently included measures of newspaper credibility such as the five-item index proposed by Meyer, but they have regularly included a question measuring confidence in the press, a both empirically and conceptually related construct. And although measures of support for free expression are largely absent from GSS surveys, the surveys have routinely posed questions based on Stouffer's classic measures of tolerance for the exercise of basic free speech rights by members of marginalized, despised or feared groups.

Free expression and tolerance are conceptually linked in that they center on forms of expression an individual may personally disapprove of but nonetheless would protect as a right in a free, open, democratic society. And, like credibility and confidence, free expression and tolerance have been shown to be empirically related.

For these reasons, the present study assumed that measures of confidence in the press and of tolerance could reasonably serve as proxy variables for measures of newspaper credibility and support for free expression.

Research Questions

Accordingly, this study used measures of press confidence, civic tolerance, newspaper reading and selected demographic variables to explore the following two research questions:


Is there appreciable evidence of associations between newspaper credibility and support for free expression among members of the American public between 1974 and 1998?


Is there appreciable evidence of associations between newspaper credibility and newspaper use among members of the American public between 1974 and 1998?

Answering these questions, it was reasoned, would contribute a valuable historical perspective to the present discussion regarding whether declines in newspaper credibility relate to declines in support for free expression and in newspaper use.


Data source and variables

The study relied on data from General Social Surveys conducted between 1974 and 1998. A GSS--provided variable accounted for oversamples of African Americans in the 1982 and 1987 surveys. Two dependent variables were considered. The first was an additive scale of civic tolerance constructed using the GSS's Stouffer-based questions.

The questions asked whether respondents would permit atheists, communists, and homosexuals to make speeches, teach in universities and circulate books through the library. The second was self-reported frequency of newspaper reading measured in five categories ranging from "never" to "every day." To aid interpretation, both measures were re-expressed as percentages of the maximum score possible. …


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.