A world of two agendas

Abstract

This article studies the efficiency of different samples for content analysis of news in media effects studies by comparing the agenda-setting effect of a classic sample with the effect of a sample drawn based on audiences’ self-reported media habits. Contrary to the belief that exposure to sampled media content is necessary for observation of media effects, samples drawn based on overall readership/viewership of the media are more efficient than samples based on audiences’ actual consumption habits. A traditional media sample yields a stronger agenda-setting effect compared to a sample drawn based on self-reported media habits. But correlations between the two media samples are also strong. The findings suggest that a broad intermedia agenda-setting process makes it possible for researchers to draw a traditional sample that is representative of the issues salient to audiences regardless of their level of exposure to the sampled media. In other words, even in a demassified media environment, traditional samples are still the best option for media effects researchers.