Comparing Coverage of Public Events Across Different Types of News Sources

University Of Wisconsin-Madison, Madison WI

Investigators

Abstract

This study uses rigorous methods to compare mainstream newswires to local specialty newspapers in their coverage of related public events in the 1990s and 2000s, an under-studied period in US history. The project simultaneously develops a catalog of such events–what happened–and a catalog of news stories that shape the impressions a reader would form about what happened. This allows demonstration of how perceptions of controversies based in mainstream national media can differ from those based in more local targeted media sources. Democracy inherently involves conflicts between people with different issue preferences, and this conflict is sometimes expressed publicly. News media shape public perceptions of controversial issues and events, but news sources differ in which issues and events they emphasize and how they cover them. While most events are mentioned only once and receive sparse news coverage, a few events are discussed in dozens of articles and become central to public discussions. Understanding these media differences is an important part of understanding democratic deliberations in a diverse society. Data about the 1990s and 2000s helps provide historical context for understanding current debates. These high-quality data will also be useful to other social scientists, and the new research methods we have developed will improve research in this area. News sources for this project are: (1) Three newswire services archived in the Annotated English Gigaword file available from the Linguistic Data Consortium; and (2) local specialty newspapers archived in Proquest Ethnic Newswatch. The main study period is 1994-2010. This study provides novel and unique data that permit enhanced investigation of the mutual relationship of public events and news coverage. Methodologically, this project develops important innovations in studying social movement events (and public events more broadly) using news sources. The methods emphasize verifiability, error-correction, and the relationality of these events. Relational databases provide rigorous links between events and the articles describing them, which both permits event data to be verified and structures the data to be consistent with theories of the interplay of events and news coverage. Consistent with theory, relational data structures are used additionally to capture the structuring of protests into specific issue clusters and the structuring of some events as complex events with subevents. These data structures permit new lines of research. Comparison across source types permits investigation of the different narratives and collective memories they construct about this period. Special attention is paid to the small number of events that receive disproportionate coverage in news sources and thus dominate public discourses. This new high-quality public event data will inform historical studies of this under-studied period and be useful for secondary analyses by other scholars. The use of relational data structures will improve data collection protocols. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.