Observing the News Flow: A Big Data Analysis of Online Coverage During the 2013 German Elections

• In Event: Methodological Advances in Digital Content Analysis: Inductive and Deductive Approaches to ‘Big Data’ Research in Journalism Studies

Abstract

While the analysis of journalistic websites seems to be a logic research path for journalism research, this route has been rarely travelled. The amount of accessible data is immense, and it is usually not available in clean form – most of the online news content is idiosyncratically structured according to the media’s own standards. Unsurprisingly, most studies that have dared to analyze online news content have done so manually, following a traditional content analysis paradigm.
This presentation will introduce a project aimed at the analysis of such unstandardized, ‘dirty’ material – namely, online news websites – using automatic steps in the retrieval, processing and analysis process.
Theoretically following an agenda building approach, the project tracked the complete news output of the most relevant news sites in Germany (42 online news sites, 28 general interest blogs) over the course of 10 weeks during the 2013 election campaign, resulting in a corpus of 265.000 news pieces. This represents most of the online news that were produced during the elections, giving access to the (nearly complete) output of professional journalism and relevant para-journalistic offerings in the form of blogs.
This allows for topological analysis (overview of the complete news flow and its logics) and a scrutiny of developments on a detail level – both at the same time.
Applying automated and semi-automated analysis steps, including NER (named entity recognition) and topic modeling, the authors extract general agenda setting principles during the campaign, including inter-media agenda setting and resonance processes in the news flow. However, they will also sketch the (current) limitations of the approach – some of them being theoretical or methodological, and some of them being technological or simply logistical.

Authors

• Elisabeth Guenther, University of Muenster

• Thorsten Quandt, University of Muenster