Page 168 - Media Effects Advances in Theory and Research

7. MASS MEDIA ATTITUDE CHANGE 157

the time were changing the attitudes of their audience, it was also possible
that the communicators were mostly attracting an audience that already
agreed with them (called “selective exposure”; see Frey, 1986) or some
combination of the two. Of course, not all analysts of the period were so
optimistic about the prospects for the mass media to produce dramatic
changes in opinion, but it was the dominant view (Wartella & Mid-
dlestadt, 1991). 1
Although the direct effects model has been replaced by more sophisti-
cated theoretical perspectives, there do remain echoes of this model
within both popular and academic writings. The news media, for exam-
ple, have been represented in the popular literature as directly influencing
and shaping political attitudes (e.g., Adams, 1993), the development of
racism (e.g., Suber, 1997), and consumer choices (e.g., Lohr, 1991). Traces
of the direct effects model can also be discerned in current theoretical per-
spectives. Zaller (1991), for instance, argues that information presentation
is the key to public opinion formation and shift. Specifically, he provides
some evidence that one can predict opinion change (e.g., attitudes toward
the Vietnam War) from the mere amount of information provided for a
particular stance (e.g., pro- or counter-U.S. involvement in the war) in the
media. As we will see shortly, most current analyses of attitude change
hold that it is not the information per se that produces persuasion, but
rather, people’s idiosyncratic reactions to this information.

Indirect Effects Model

The direct effects model was tempered considerably in the next two
decades, largely as a result of the subsequent empirical research con-
ducted. For example, in analyzing survey information gathered by the
National Opinion Research Center, Hyman and Sheatsley (1947) con-
cluded that the effectiveness of mass communication campaigns could
not be increased simply by increasing the number of messages. Rather, the
specific psychological barriers to effective information dissemination
must be considered and overcome (see also Cartwright, 1949). For exam-
ple, they noted that people often distort incoming information to be con-
sistent with prior attitudes, making change less likely. A similar conclu-
sion was reached by Lazarsfeld, Berelson, and Gaudet (1948) in their

1 In one of the relatively rare empirical efforts of the period, Peterson and Thurstone (1933)
examined the power of movies such as D. W. Griffith’s Birth of a Nation, controversial because
of its depiction of Blacks, to modify the racial attitudes of adolescents. The conclusions of this
research foreshadowed the modern period in that various moderators of effective influence
were uncovered (e.g., greater influence for those with low knowledge rather than high issue-
consistent knowledge; Wood, Rhodes, & Biek, 1995; see Wartella & Reeves, 1985).