Page 92 - Media Effects Advances in Theory and Research

4. MEDIA CONSUMPTION AND PERCEPTIONS OF SOCIAL REALITY 81

tic. Rather, the perception of ease may be the causal mechanism, which
may or may not be adequately captured by response time (see also
Schwarz et al., 1991, regarding issues of ease of recall).

Proposition 2: Accessibility Mediates the Cultivation Effect. Propo-
sition 1 (viewing influences accessibility) is a necessary but not sufficient
condition to implicate the availability heuristic as an explanation for culti-
vation effects. It is also necessary to demonstrate that accessibility mediates
the relation between level of viewing and magnitude of judgments (Manis
et al., 1993); that is, it is also necessary to demonstrate that the enhanced
accessibility leads to higher estimates. Otherwise, it could be argued that
television viewing impacts accessibility and the magnitude of the judg-
ments independently.
Some indirect evidence of the mediating role of accessibility was pro-
vided by Shrum and O’Guinn (1993). When accessibility (speed of
response) was controlled, the cultivation effect was, for the most part,
reduced to nonsignificance. More direct evidence of mediation was pro-
vided by Shrum (1996). Following the procedure used by Manis et al.
(1993), path analyses were used to demonstrate that the level of television
viewing was related to accessibility (again, operationalized as response
latencies), which in turn was related to the magnitude of the estimates.
However, the path analyses also revealed that the mediation was only a
partial one: Television viewing still had a direct effect on the magnitude of
the estimates, even when the influence of accessibility was controlled.
Busselle (2001) also provided evidence of the mediating role of accessi-
bility by manipulating the conditions under which the prevalence esti-
mates for particular constructs (e.g., a shooting) were constructed. Some
participants provided their prevalence estimates before recalling an
example of the construct (judgment-first condition) whereas other partici-
pants recalled an example before providing their estimates (recall-first
condition). Level of television viewing was expected to make an example
easier to recall in the judgment-first condition, whereas recalling an exam-
ple before judgment was expected to make an example equally accessible
for all participants, regardless of television viewing level. The results con-
firmed these expectations.

Proposition 3: Television Exemplars Are Not Discounted. An implicit
assumption in the notion that the availability heuristic can explain culti-
vation effects is that the examples that are retrieved and used as a basis for
judgment are considered relevant or applicable to the judgment. This is an
important assumption because research has shown that accessibility
effects typically obtain only when this condition is met (Higgins, 1996).