Page 287 - Media Effects Advances in Theory and Research

276 SPARKS AND SPARKS

not increase in South Africa from 1945 to 1974 while a ban existed on TV.
However, as soon as this ban was lifted, homicides began to increase there
as well—more than doubling in less than 20 years, just as it had in the
United States and Canada. Centerwall concludes that the data he exam-
ined indicated that about half of all homicides in the United States are
caused, in part, by exposure to TV.
Phillips has also analyzed naturally occuring data and reached conclu-
sions that are similar to Centerwall’s. With respect to homicides, Phillips
argues that after widely publicized heavyweight prize fights, the homi-
cide rate increased. Similarly, he notes that after news stories of widely
publicized suicides, increases occurred in single-car fatalities and airplane
crashes. Of course, unequivocal conclusions about causality are not possi-
ble based on the type of data presented by Centerwall and Phillips. How-
ever, other researchers have recently analyzed their own data and drawn
conclusions that support the validity of the claims made about the link
between media violence and subsequent homicides and suicides (Cantor,
Sheehan, Alpers, & Mullen, 1999).

Surveys

As in natural experiments such as the ones reported by Phillips and Cen-
terwall, surveys on the topic of media violence and aggression are
designed to add data that are free from the constraints of the laboratory. In
exchange for this freedom, researchers who use the survey methodology
give up the ability to make conclusive claims about causality. Although a
few surveys report little or no relationship between media violence and
aggression (Milavsky, Kessler, Stipp, & Rubens, 1982; Singer & Singer,
1980), the most ambitious survey research that has examined the relation-
ship over a period of decades (Huesmann & Eron, 1986) seems to draw
the same general conclusion that emerges from experimental studies.
These researchers collected data from children when they were 8 years old
and followed these same children in a panel study until they were
30 years old. Those who watched the highest levels of TV violence as chil-
dren were more likely to be involved in serious crime when they were
adults. Huesmann (1986) summarized the basic conclusion from this
work by stating: “Aggressive habits seem to be learned early in life, and
once established, are resistant to change and predictive of serious adult
antisocial behavior. If a child’s observation of media violence promotes
the learning of aggressive habits, it can have harmful lifelong conse-
quences. Consistent with this theory, early television habits are in fact cor-
related with adult criminality” (pp. 129–130) (see also Huesmann, Moise,
& Podolski, 1997).