Many to some degree get closer to identifying

Many studies to date have examined whether “playing
violent video games would heighten the level of aggression
or increase aggressive behavior” 3, 5. Previous studies
can be roughly divided into experimental studies, survey
studies, and panel studies, and they often indicated the
possibility that video games could have adverse effects on
aggression or aggressive behavior. In laboratory
experiments, those who are in the experimental group play
aggressive video games in a laboratory, and their aggression
is later compared with that of the control group. In general,
laboratory experiments have a strong ability to identify
causal relationships; however, they have problems in that
experimental scenes are artificial and that identified causal
relationships are usually short-term relationships.
Meanwhile, in survey studies, the amount of video game
use and aggression of the participants are measured to
examine the correlation between them. Although some of
the previous studies indicated that there was a significant
positive correlation between the amount of video game use
and aggression 2, 11, 7, one-wave studies cannot identify
the cause and effect even though they can identify
correlations. Survey studies, therefore, overcome the
artificiality problem, but are still unable to identify causal
relationships.
Although it is still a survey study, a panel study
can to some degree get closer to identifying causal
relationships than other methods that obtained correlations
in one-time studies 4. A panel study is a survey study in
which the same variables are measured multiple times with
the same subjects. Causal relationships can be estimated by
analyzing data obtained through this method. There have
not been many panel studies so far; therefore, we conducted
a panel study in order to examine the effect of video games
on aggressive behavior in real life.
So far, the amount of time spent on using video
games and games of specific genres were used as
independent variables in panel studies8, 6. In addition to
these variables, in this study, we measured the amount of
time that the participants were exposed to violent scenes as
well as their preference for violent games. Although the
amount of video game use changes with the situations that
the participants are in, preference usually stays the same;
therefore preference has the advantage that it remains stable
in terms of time.
We also studied not only aggressive behavior but
also pro-social behavior as dependent variables. A negative
relationship between violent video games and pro-social
behavior has been reported 1. One of the examples in
which aggressive behavior and pro-social behavior were
used as dependent variables in a study is the study by
Wiegman & van Schie (1998) 10. They used the amount
of use and preference for violent games as video game
variables, and examined correlations between the video
game variables and dependent variables. As a result, they
suggested that there was a significant negative correlation
between the amount of video game use and pro-social
behavior, and that boys exhibiting strong preference for
violent games were likely to show aggressive behavior and
unlikely to show pro-social behavior compared to those
exhibiting weak preference for violent games. In this study,
170
we conducted a panel study to examine if playing violent
video games would cause an increase in aggressive
behavior and a decrease in pro-social behavior