“Men prefer redheads”: Media framing of polls and its effect on trust in media

How are polls presented in the media, and how does their presentation affect trust in the media? Whereas publishing surveys and interpreting scientific results can advance an image of precision and neutrality, the media publish polls of varying methodological quality. The same newspaper can publish rigorous (“objective” seeming) election studies, as well as unrepresentative polls regarding sexual habits (“subjective” polls). We argue that to communicate the difference among surveys to consumers, the media create different frames for different surveys, each characterized by a distinctive presentation pattern (e.g., inclusion of diagrams, size and placement of article, extent to which methodology is explained, etc.). Here, four such frames and presentation patterns are suggested, corresponding to four media role perceptions, and generally divided into frames presenting the poll as “objective” versus “subjective.” These frames were examined using a content analysis of a sample of articles with surveys published in the Israeli daily Yedioth Ahronot and confirmed using a validating experiment. We also believe that subjective polls reinforce perceptions that the media are unprofessional and inaccurate, manifesting in cynicism and distrust. An experiment using random assignment indeed demonstrates that exposure to “subjective” polls decreases trust in the media, the implications of which are discussed.