CONTENT STUDIES OF MEDIA REPRESENTATIONS
Some questions about generalities can be answered in relatively precise
terms by measuring or counting certain textual features. Since the 1940s, such
quantitative analyses have played an important role in several forms of media
research. In the first general account of this type of analysis, Bernard Berelson
(1952) used the term ‘content analysis,’ defining it as a research technique ‘for the
objective, systematic, and quantitative description of the manifest content of
communication’ (p. 18).
As this definition suggests, content analysis is basically descriptive. Certain
well-defined textual elements or characteristics are measured in various ways;
for example, the amount of newspaper coverage of certain issues, the topics of
television news programs, the occurrence of particular symbols in fictional
genres, and so on. Such quantitative descriptions of the features of a group of texts
are then typically used as a basis for more general inferences about the meaning
of these texts and their implications regarding various social phenomena.
One early example of this approach was Leo Lowenthal’s study of ‘mass
idols’ from 1941. In a quantitative analysis of biographical articles in popular U.S.
magazines, he showed that while political leaders, businessmen, and scientists
were portrayed in 46 percent of the cases at the beginning of the twentieth
century, this percentage had dropped to 25 by 1940, and the interest was now
focused, instead, on film actors, entertainers, and athletes. This shift had begun
in the 1920s, and was accompanied by an additional shift of focus from the public
life and achievements of the people in question, to their love life, their ways of
dressing, and other private matters. Lowenthal interpreted this change of
emphasis from ‘idols of production’ to ‘idols of consumption,’ and from the public
to the private sphere, as a textual expression of general changes in social values
during the period.
Since the early days of content analysis, then, this quantitative methodology
has been used in a variety of research contexts, often in combination with other
methodologies. The main varieties have included (for a more detailed overview
see Wimmer and Dominick 1994):
1. Studies of patterns and trends in media representations as such;
2. Studies of the relationships between textual representations and ‘the real
world’;
3. Studies concerning media effects.
In some studies, the overall content of a specific media genre (e.g., television
news) has been analyzed by the Glasgow University Media Group (1976 and
1980). Other studies have undertaken analyses of how the media handle specific
events, such as political demonstrations (e.g., Halloran et al. 1970) and wars (e.g.,
Morrison and Tumber (1988) on the Falklands conflict, and Morrison (1992) on
the Gulf War).
The depiction of violence, particularly in television fiction, has been a favorite
topic. In such studies, researchers commonly start by defining how ‘violence’
should be understood, and in which specific discursive contexts it should be
measured. Next, coders watch samples of television programs and count the
incidents which match the definition. The longest running and most well known of
these studies was started in the late 1960s by George Gerbner. In the initial
research design, he defined violence as ‘the overt expression of physical force
(with or without a weapon) against self or other, compelling action against one’s
will on pain of being hurt or killed, or actually hurting or killing’ (Gerbner 1972: 31).
From this definition, Gerbner further established a series of guidelines regarding
the types of events to be included in studies with reference to their place in the
fictional universe and their relation to the overall plot (see Gerbner and Gross
1976; Gerbner et al. 1977). Continued during the 1990s by Gerbner and his team,
this series of studies is an example of research projects in which content analysis
is used to compare media representations with actual occurrences in society.
These studies have repeatedly shown that the world of U.S. primetime television
fiction is a far more violent place than the real world, and that the portrayal of
certain social groups as frequent victims of crime is inconsistent with social
statistics. The same set of studies has also addressed media effects. Comparing
content analyses with survey data, Gerbner and his team have argued that
television fiction ‘cultivates’ or shapes people’s perceptions of social reality, so
that heavy viewers tend to have misconceptions of the role of violence in everyday
life.
