Page 99 - Media Effects Advances in Theory and Research

88 SHRUM

Cacioppo, 1986). In fact, given that the model shown in Fig. 4.3 derives
from the concepts of heuristic (peripheral) and systematic (central) pro-
cessing, it is no accident that the model bears striking similarity to the
ELM (as well as the Heuristic/Systematic Model, Chaiken et al., 1989). In
general, the less the elaboration (due to such things as time pressure, men-
tal ability, involvement, etc.), the greater the cultivation effect.

Implausible Alternative Hypotheses

Although some, if not all, of the studies that have formed the basis of the
model have potential alternative explanations, it is difficult for concepts
such as spuriousness or reverse causality to account for the general pat-
tern. For example, the initial studies that tested Propositions 1 and 2
(accessibility) were strictly correlational and thus could be explained in
terms of either spuriousness or reverse causality. However, these alterna-
tive explanations cannot account for the results of the experiments that
tested Propositions 3 through 5, particularly the pattern showing that
both the experimental manipulations of source priming and of processing
strategy produced nearly identical results, with the manipulations reduc-
ing estimates of heavy viewers to the equivalent of light viewers, but leav-
ing the estimates of light viewers unaffected.
The consistency of results across the different types of dependent vari-
ables also argues against explanations other than a causal effect of televi-
sion viewing. Consistent results tended to be found for judgments of
occupational prevalence (doctors, lawyers, police officers), crime, and
affluence (and, to a lesser extent, marital discord). Although reverse
causality or spuriousness explanations can be used (and often are) to
explain the results for any one variable, it is difficult to account for the
effects on all variables. Rather, the more parsimonious explanation is that
the causal factor is the one that they most have in common: They are con-
structs overrepresented in television portrayals relative to their real-
world incidence.

Reconciling Conflicting Findings in Past Research

As mentioned earlier, one of the useful features of a process model for cul-
tivation effects is that it has the potential to reconcile conflicting findings
that have been reported. The myriad of paths toward little or no cultiva-
tion effect that are shown in Fig. 4.3 has the potential to explain some of
these conflicts.

Source-Priming Explanations. The source-priming manipulation
used by Shrum et al. (1998) had participants provide information on how