Page 263 - Culture Society and the Media

CULTURE, SOCIETY AND THE MEDIA 253
been controlled for the educational level of the respondents. Such control is
necessary, because the association of mistrust of politicians with dependence on
television could have been due simply to the fact that the less educated
respondents were both more prone to mistrust and heavy users of television. But
the bottom part of the Table suggests that mistrust and dependence on television
went together even when educational level was held constant.
On the face of it, this pattern seems clear and its interpretation
straightforward. Nevertheless, its methodological foundations were attacked by
Miller, Erbring and Goldenberg (1976) on two grounds. First, they asked, where
is the crucial evidence demonstrating the negative and antiinstitutional quality of
television news, which is supposedly productive of viewer mistrust? Without
some analysis of the news content of television, they maintained, a vital link in
the supposed chain of causation is missing. Second, they argued, subjective
statements about reliance on a medium provide no measure of an individual’s
actual exposure to the medium or to that part of its contents that supposedly act as
a trigger of mistrust. Controlling for educational level was also, according to
these critics, insufficient to rule out the influence of other factors, since other
studies have shown dependence on television to go with low levels of political
information and interest, independently of levels of education, and those are
characteristics that could be related to cynicism and mistrust. Then, as if to clinch
these arguments, they examined the relationship between frequency of television
news viewing and attitudes similar to those examined by Robinson, and reported
little evidence of any effect.
These authors also presented some evidence of their own on the same subject,
striving to apply their methodological prescriptions in practice. For this purpose,
they looked at change over time, examining data taken from members of a
national sample who had been interviewed first in 1972 and again in 1974 and
who had on both occasions responded to five statements comprising a so-called
trust in government scale. Within the sample, they compared the readers of two
kinds of newspapers—whose front-page stories presented, as ascertained by
content analysis, above-average and below-average amounts of criticism of
politicians or political institutions—to see whether members of the former group
displayed a bigger increase in mistrust of government. Their data suggested that
mistrust had in fact increased over the two-year period throughout the American
public—even among readers of relatively uncritical newspapers. They also found
that the tide of increasing mistrust reached higher levels among those with
lower levels of formal education. However, the growth of mistrust over time was
greater among readers of critical papers than among readers of the uncritical
ones. The differential widened considerably when the analysis was confined to
frequent readers of the two sorts of newspapers. It looked as if frequent readers of
the more critical papers were especially receptive to their papers’ critical outlook.
The authors concluded their analysis on the following note: