This paper discusses what contributions quantitative methods （roughly
equal to statistical methods） have made to media research, especially to traditional
mass communication research. Keeping in mind that statistical methods
and case-study methods are not mutually exclusive, this paper points out a
threefold merit of the former. First, statistical methods enable researchers to
infer how widely various uses and effects of the media are distributed in the
population（ e.g., the uses and gratifications study from the 1970s onward）. Second,
easier replicability of theoretical hypotheses has made for cumulative
advancement of a specific research area （e.g., the agenda-setting effect
research）. Third, counterintuitive findings sometimes found in empirical data
would bring an opportunity for coming up with a new theory（ e.g., the two-step
flow of communication model）. Conceptual models （or hypotheses） that the
empirical studies of mass communication have advanced and cumulated so far
are defined as middle-range theories Sociologist Robert K. Merton once advocated.
These models are common properties for media researchers. Even if
some of the models might become less relevant in a new media landscape in
the 21st century, we should seek to make the best use of these properties to
“see further on the shoulders of giants.”
