Page 81 - Media Effects Advances in Theory and Research

70 SHRUM

these relations (Hawkins & Pingree, 1990; Reeves, Chaffee, & Tims, 1982;
see Wyer, 1980, for a similar view on social psychological research).
Although the purpose of this chapter is to address the criticism per-
taining to the lack of a cognitive process explanation for observed media
effects, the two criticisms just noted are not independent. One of the use-
ful features of process explanations is that ideally models are developed
that can specify both moderating and mediating variables. McGuire
(1986) notes in his fairly exhaustive review that even though research to
date has shown remarkably small media effects, there are a number of
possibilities that may ultimately allow for the “salvaging” of the massive
effects notion. In particular, he notes that small main effects may be
obscured by messages having different effects on different groups or as a
function of different situations (moderators) and by focusing on direct
effects at the expense of indirect ones (mediators). Thus, the development
of cognitive process models for media effects has the potential to uncover
new relations as well as to make sense out of old ones.
The development of cognitive process models that can explain media
effects has other advantages as well. For one, it has the potential to
increase internal validity, or the extent to which we are confident that we
are observing a true causal effect and not one that is spurious (Hawkins &
Pingree, 1990), another common criticism of many media effects studies
(see Hirsch, 1980; Hughes, 1980; McGuire, 1986). A process model should
provide clear links between the stimulus (e.g., media consumption) and
the response (e.g., beliefs, behavior), and each link in the model should
represent a testable proposition to be empirically verified. If these links
stand on solid theoretical foundations and are empirically verified, then
threats to internal validity such as spuriousness and reverse causality are
rendered less plausible, as the threats would presumably have to occur at
each stage. Another advantage is that process models may potentially
address conflicting findings in previous research. A process model should
provide boundary conditions for the effect; that is, a specification of the
conditions under which the effect does not hold. To the extent that these
boundary conditions are related to aspects of inconsistencies in previous
research, disparate findings may be reconciled.
Given these advantages of a focus on process, the goals for this chap-
ter are twofold: (1) to discuss some of the general underlying principles
that have emerged in social cognition research that have particular impli-
cations for media effects, with reference to relevant media effects re-
search that exemplify these principles, and (2) to demonstrate the advan-
tages of a process focus just discussed by outlining a cognitive process
model that can explain a particular media effect, the cultivation effect
(see chapter 3).