Page 95 - Media Effects Advances in Theory and Research

84 SHRUM

process systematically when constructing their judgments. Compared to
heuristic processing, systematic processing is associated with the con-
sideration of more information and greater scrutiny of the information
that is considered. Systematic processing is used when it is important to
determine the validity of information (Petty & Cacioppo, 1986) and has
been shown to attenuate the effects of heuristics (Chaiken et al., 1989).
Under systematic conditions, it seems likely that the relation between
level of viewing and social perceptions would be weakened or elimi-
nated entirely. When people process systematically, they should be more
likely to retrieve examples other than simply the first ones that come to
mind, should be more likely to scrutinize the retrieved information, and
thus should be more likely to ascertain and discount information from
unreliable sources such as television programs, than when they process
heuristically.
One condition that is related to whether heuristic or systematic pro-
cessing strategies are adopted is the motivation to process information
(Sherman & Corty, 1984): When motivation is high, systematic processing
predominates; when motivation is low, heuristic processing predomi-
nates. Moreover, motivation is itself determined by a number of factors,
including level of issue involvement (Petty & Cacioppo, 1990) and level of
task involvement (Chaiken & Maheswaran, 1994).
To test Proposition 4, Shrum (2001) manipulated the processing strate-
gies that participants used to construct their estimates of the prevalence of
crime, marital discord, affluence, and certain occupations. Some partici-
pants were induced to process systematically via an accuracy motiva-
tion/task importance manipulation (Chaiken & Maheswaran, 1994), oth-
ers were induced to process heuristically by asking them to give the first
answer that came to mind, and a third (control) group received no manip-
ulation, but were simply instructed to provide their estimates. Television
viewing was then measured after the judgments were made. The results
were consistent and as expected. Both the control group and the heuristic
group produced cultivation effects that did not differ in magnitude from
each other. However, the systematic group showed no cultivation effect.
Moreover, the pattern of results was remarkably similar to those obtained
by Shrum et al. (1998, Study 1): The estimates of light viewers did not dif-
fer as a function of condition, but the systematic condition affected only
heavy viewers, bringing their estimates more in line with those of all light
viewers, regardless of processing condition. This pattern of results can be
seen in Fig. 4.2. 4

4 As with Fig. 4.1, the graph shown in Fig. 4.2 represents the general pattern of results
across dependent variables. For details of the actual effects for each dependent variable, see
Shrum (2001).