'To follow or not to follow?' The Twitter network of Belgian health journalists

This paper puts forward two main objectives. First, it sets out to elaborate the academic debate on the news sourcing practices of professional journalists in the network society. Second, it contributes to the methodological development of journalism studies in exploring the added value of innovative research methods, more specifically digital methods and social network analysis. The increased speed of information dissemination, the interactivity and the connectivity that characterize social media sites offer the potential for journalists to revise their sourcing practices in light of the often-cited dangers of heavy reliance on elite sources such as government actors, PR sources or experts (Castells, 2008, 2011; Gans, 1979, 2011; Heinrich, 2011). Especially Twitter has proven to be the preferred social media news gathering channel for professional journalists due to its characteristics that invite information sharing (Hughes & Palen, 2009). Yet the question remains whether journalists make full use of the new possibilities available through social media. We build on two contrasting views that dominate the sourcing debate. On the one hand, there is the bottom-up perspective suggesting that social media channels allow users to spread information cheaply and instantaneously throughout their network. As a result, they can open the news gates for non-elite sources (Kaplan & Haenlein, 2010). On the other hand, several authors conclude that journalists stick with their old top-down sourcing routines and continue to privilege elite sources, also through social media channels (Author, 2013). Indeed, several studies point out that social media channels are “colonized” by elite sources, who understand the potential of social media to strengthen relations with journalists (Broersma & Graham, 2012). With this paper we want to contribute to this academic debate and clarify the sourcing practices of professional journalists. We examined the Twitter sources of Belgian health journalists by means of social network analysis and digital methods. We focus on Belgian health journalists for different reasons such as the topicality of the health care debate and the -defined and therefore feasible- network of 6 Belgian journalists specialized in health news and having an active Twitter account. First, we used the Twitter REST API and the R package twitteR to collect all “follows” of each journalist. These “follows” can be considered as sources since journalists generally follow actors because they expect them to share newsworthy information. This resulted in an automatically generated list of 1530 “follows” of our 6 journalists, or 1220 unique Twitter profiles (some sources are followed by more than one journalist). Each “follow” or source was then manually coded in terms of ten different categories or “identity attributes” (e.g. health insurance company, patient organization, politician, pharmaceutical company, expert) and in terms of relatedness to the health or scientific field. Descriptive results confirm Rupar’s (2015) finding of “Twitter as a press club”. 41.1 percent of all “follows” are media sources. Twitter is thus mainly used by our 6 health journalists to engage with other journalists and media outlets. The second most important group of sources are experts such as scientists or medical doctors (18.7%) which is in accordance with the idea that health information is often very complex which makes journalists turn to experts on the matter for clarification and contextualization (Len-Rios et al., 2009). Twitter is not intentionally used by the journalists to reach out to citizens. Only 4.7% of the “follows” are ordinary citizens. 38.4 percent or 468 of the 1220 “follows” are related to the health or scientific field. If we only take this subgroup into account which is most valuable for health information, we see that experts are the most important sources (46.6%) followed by media sources (17.1%). The number of ordinary citizens decreases to 1.3%. In other words, our findings indicate that sourcing practices of health journalists on Twitter resemble their offline sourcing practices. Journalists follow elite sources, especially media sources and experts, while non-elites remain largely invisible. In a next step of the study, which will be executed in December 2015, we will reconstruct the ego-network of each health journalist and then aggregate the six networks to get insight into the overall “following” network of our 6 Belgian health journalists. Sources who are followed by several journalists take a more central position in the overarching Twitter network of Belgian health journalists, which will give us an indication of their importance as a news source