ABSTRACT
Problems with that notion soon became apparent. Researchers had trouble identifying who were opinion leaders and who were not. Opinion leadership differed from topic to topic, and designating a particular person within a social group as the opinion leader became difficult. Furthermore, data from the original study by Katz and Lazarsfeld (1955) indicated that most of the interpersonal discussions generated by the media were horizontal in nature-between social equals-rather than a downward transmission from leaders to followers. As a result, researchers in opinion leadership soon shifted their focus to an analysis of social influence produced by those horizontal interpersonal discussions.
