29
burglaries, larceny theft, aggravated assault, murder, forcible rape, property crime, or
motor vehicle theft and the amount of violence in any of the four media types. Thus, the
answer to the question posed in RQ2, does presence of violence in the media relate to
crime data, appears to be no.
Next, bivariate correlations were run between the violent content variables and the
numbers of times that the issue of media violence was covered in the news media, using a
Lexis Nexis search. The latter set of variables included news coverage of the phrases
“media violence,” “movie (or film) violence,” and “television violence” and included
searches of Lexis Nexis categories major newspapers, newsmagazines, and television
news. Results show that only two of such news coverage of media violence variables
were correlated with amount of violent content. First, the number of Lexis Nexis hits in
major newspapers on the topic of movies and violence was significantly and positively
correlated with the percentage of the total media content examined that contained
violence (r = .43, p < .05). Second, the number of Lexis Nexis hits in newsmagazines for
the same subject, movies and violence, was also significantly and positively correlated
with percentage of violent content (r = .58, p < .01). Therefore, in response to RQ3, there
is only scant evidence of a connection between the news media attention to the issue of
media violence and the actual amount of violence in the media.
Finally, additional analyses were conducted to investigate RQ4 by determining
whether the amount of violent content in the media fluctuates in association with
presidential and congressional attention to the issue of media violence. Of all of the seven
variables that were used to determine indicators of concern on the part of federal policy
makers about the issue of media violence, only one correlated significantly with only one
Convention