Eirik Stavelin

ABSTRACT This article presents a longitudinal big data analysis of the online news provision of the Norwegian Broadcasting Corporation (NRK). Comparing data from 2009 with data from 2013 (approximately 150,000 news items), the analysis assumes a hybrid methodological approach combining human and computer coding to enquire about longitudinal trends in online journalism as produced by a public service institution in a competitive market. The analysis finds NRK's online news profile to be rather stable across the period, with increases in the areas of audiovisual streaming, linking practices, multimedia features and social media sharing as a result of technological developments. While a more dynamic front-page environment puts the broadcaster as a competitor in the overall online news market, an increase in self-promotional content also suggests NRK is moving towards a more pronounced broadcasting identity. As public service broadcasters are being increasingly challenged across Europe by claims of unfair competitive advantage, this analysis offers a much-needed empirical basis for discussing the role and function of public service broadcasting online.

This chapter describes a new method for analyzing online news employing both human and computer coding and based on a triangulation of qualitative and quantitative measures. The chapter illustrates how the method was used to perform a content analysis of the full online news output of the Norwegian Broadcasting Corporation (NRK) for the year 2009 (n = 74430). It explains how the approach was developed, describes the problems we encountered along the way, and provides a rationale for further use of the method. We conclude that our work demonstrates that methodologies designed for measuring broadcast news do not suffice in the ephemeral online news environment. Thus, online research methods need to be redesigned to account for medium-specific news features on the Internet.

ABSTRACT This article examines computational journalism as a craft practised in Norwegian newsrooms. Based on in-depth interviews with expert practitioners in six of the largest newsrooms in Norway, we find that computational journalism represents a continuation of traditional (investigative) journalism. While the skills and tools necessary to do this kind of journalism diverge from the typical journalist's, the values and aims align well with tradition. Even though computation enables journalists to cope with the size and scale of journalistically appealing datasets, we find little evidence for computational journalism to increase the efficiency of doing journalism or in any other way rid journalists from low-level technical work.