The impact of expertise and affect on political information processing.

Abstract

This dissertation investigated the ways in which individuals' processing of political information may be jointly influenced by their political expertise and current affective states. In particular, an experiment was conducted to examine the various conditions under which individuals' political expertise and the valence of their affective states (positive or negative feelings) might, independently and in interaction with each other, shape the cognitive strategies involved in processing political news. Two hundred and six Michigan undergraduates participated in the experiment. Subjects were randomly assigned to read a set of initial news stories to induce either a positive or negative mood. After manipulation checks, subjects read a prepared newspaper article, either on a change in Michigan's student loan program (low-salience issue), or on the deputization of campus police (high-salience issue). Subjects were then asked to list all thoughts that occurred to them while reading the story. Following the thought-listing exercise were factual questions about the issue, to measure subjects' domain-specific expertise, and questions pertaining to global political affairs in order to measure general political expertise. Major dependent variables included the total number of thoughts generated by subjects, the number of dimensions used, the proportion of thoughts classified as issue-relevant, and the proportion of thoughts classified as arguments. In line with prior research, political expertise emerged as an important predictor of the degree to which people engaged in analytic processing of the news. The respective impact of general political expertise and domain-specific expertise varied somewhat, depending upon the specific cognitive outcome being observed. The impact also varied by the salience of the issue. For the high-salience, but not the low-salience issue, general political expertise generally failed to predict analytic processing. The hypothesis that negative affect would induce more analytic processing received only very weak support. For both high- and low- salience issues, an interaction between political expertise and affective state was observed, supporting the hypothesis that expertise and affect operate in combination to influence analytic processing. But, contrary to initial predictions, negative affect induced political experts to exert more use of analytic strategies, while it decreased novices' analytic processing. The interaction was more pronounced with the low-salience, student loan issue. The generalizability of the findings are discussed, along with implications of the experimental results for mass communication research, particularly in the areas of media exposure, agenda-setting, media priming, and negative political advertising.