Page 121 - Media Effects Advances in Theory and Research

110 D. ROSKOS-EWOLDSEN, B. ROSKOS-EWOLDSEN, F. DILLMAN CARPENTIER

Furthermore, despite the commonalities across these theories, their
domains differ too much to afford a single theory of media priming. For
example, the affective aggression model’s (Anderson et al., 1995) reliance
on network models for explaining affective priming is problematic
because recent research has seriously questioned the ability of network
models to explain affective priming (Franks, Roskos-Ewoldsen, Bilbrey, &
Roskos-Ewoldsen, 1999; Klinger, Burton, & Pitts, 2000). In addition, a
unique feature of this model is that it incorporates secondary appraisals
that can override the effect of the priming events on subsequent behavior.
Clearly, this is a necessary addition to the model because it allows the
model to explain how the priming of aggressive cognitions and affect
does not always result in aggressive behavior. However, it is unclear how
this component of the model would apply to political priming. Con-
versely, in Price and Tewksbury’s (1997) model, judgments of applicabil-
ity play a central role in determining whether activated nodes serve as
primes or whether they influence how a media story is framed. Recall that
judgments of applicability involve whether the particular construct that is
activated by the media is applicable to what is being watched/read.
When an activated construct is judged as applicable, it is then used to
frame what is being watched/read and does not act as a prime. When it is
judged as not applicable, then it can serve as a prime. In the context of
media violence, the violence, although often not necessary, is typically
applicable to the show. As a consequence, the violence would trigger a
judgment of applicability, which should afford no aggressive priming by
the media. Obviously, this is not the case. Ultimately, the difficulty in com-
bining the different models of media priming into a coherent model that
covers all the domains is that the current models were specifically crafted
to explain the findings relevant to only that domain of study.
In our view, network models of media priming provide a starting point
for understanding the effects of the media on subsequent judgments and
behavior. However, we believe that network models need to be subsumed
within a larger theoretical framework to explain adequately the phenom-
ena that these models are attempting to explain. Following, we propose
such a theoretical framework, a mental models approach.

A NEW FRAMEWORK FOR UNDERSTANDING MEDIA PRIMING:
THE MENTAL MODELS APPROACH

The mental models approach reflects the observation that thinking typi-
cally occurs within and about situations (Garnham, 1997). Mental models
are the cognitive representations of situations in real or imaginary worlds
(including space and time), the entities found in the situation (and the