Testing the Cognitive Mediation Model

Abstract

With data from a 2000 telephone survey of a Midwestern community, the current study tests and reevaluates the cognitive mediation model. In doing so, the authors experiment with a news reliance index and three gratifications sought dimensions: surveillance, anticipated interaction, and guidance. Although there is support for the surveillance and guidance versions of the cognitive mediation model, findings change greatly with the inclusion of anticipated interaction gratifications sought. In a model with the three gratifications sought, the effects of surveillance and guidance on political knowledge fall out, whereas that of anticipated interaction is direct and unmediated. The authors explain these findings with reference to previous research and comment on how different measures of media use and gratifications sought may alter the cognitive mediation model.