ABSTRACT
Comparative communication research has become a growth field in terms of the amount and complexity of its scholarship. This was not always the case. From its origins in the 1950s, most empirical communication research in the United States ignored structural variations in concentrating on individual behavior. The authors of the classic Columbia University voting studies consciously avoided comparing their research in small homogeneous communities with the many whole-community studies of that era in order to establish “universal propositions generalizable across time and space” (Berelson, Lazarsfeld, & McPhee, 1954). This strategy produced considerable knowledge about communication processes at the individual level. Unfortunately, the focus on universals relegated social structural differences as sources of influence to be removed by statistical control. It inadvertently diverted attention from important sources of differences between communities that might have led to the development of more nuanced multilevel macro theories.
