Blindfolding the public : examining the hydraulic pattern hypothesis of media priming effects

In this dissertation was examined the hydraulic pattern of media-priming effects by looking into Granger causality (a statistical test to determine if one time series is useful in forecasting another) between media coverage and the importance of issues people perceive. The hydraulic pattern hypothesis, an argument that increase in the importance of an issue is accompanied by decrease in a similar amount of importance, is embedded in most media-effect theories but has rarely been tested. To test the causality with media coverage, time series of six issues and six candidate variables were created. This research is distinct from previous studies of priming in that it tests aggregate-level influence of media coverage on popular evaluation of political-campaign candidate in a long-term setting. In the findings, media coverage of issues induced changes in the Granger-caused issue-weight of the issue that it covered, confirming the main effects of priming. The hydraulic pattern was also confirmed. Active media coverage of an issue, induced Granger-caused changes in five other issue-weights. It was found that it takes 7–8 days after the media coverage to establish a causal relationship of priming effects. vii In another finding, the result showed that the time-lag of the hydraulic pattern preceded the main priming effects. As regards the debated relationship of priming effects with political knowledge, this research found that high knowledge groups are more susceptible to the main priming effects. However, the impact of political knowledge on the hydraulic pattern was the opposite. This means that less knowledgeable people may be more vulnerable; that is, they are more likely to lose sight of other issues when the media primes a certain issue. In the test of attribute priming, the causality of the hydraulic pattern was also established to a lesser degree. Especially, personality-related candidate attributes like trustworthiness were robust regarding the hydraulic-pattern effects. In all of these analyses, the measurement of optimal time-lag was utilized instead of the durability concept used in previous studies. With this study design and new measurements, this research contributes to the literature by providing new insight into the theoretical conundrums related to priming theory. One of such insight is that the priming effects that matter at the poll, are relatively slow and deliberative processes, and are differentiated from the temperamental daily effects of news.