Opinionet | Online media networks and audience flow study at AECPA Congress in San Sebastián
post-template-default,single,single-post,postid-29,single-format-standard,ajax_fade,page_not_loaded,,wpb-js-composer js-comp-ver-4.1.2,vc_responsive


Online media networks and audience flow study at AECPA Congress in San Sebastián

  |   Uncategorized

This year for the first time, the Spanish Association of Political Science (AECPA)  organized a panel focused on digital methods during its annual Congress. The event took place in San Sebastian last July and communications assessing a broad range of political sciences questions were invited.  Our current investigation on online media networks and audience flow which address the level of fragmentation in news production and consumption on the Spanish web, was accepted to be presented during the Congress.


PhD Candidate Sílvia Majó-Vázquez explained the preliminary results to the researchers and political scientist attending the event in Euskal Herriko Unibertsitatea (Univesidad del País Vasco). The study, co-authored by professors Ana S. Cardenal and Sandra González-Bailón, offers a comprehensive analysis of the Spanish online public sphere by mapping its two fundamental networks: Firstly, the network of hyperlinks among news media; and secondly, the audience network built by tracking people when navigating political content. By means of the two structures, the study unveils the drivers that influence modes of exposure to political information. It applies tools from network science to operationalize fragmentation in the online domain. In particular, it identifies the community structure of the news media network using a random walker approach, a method that helps to identify network areas with denser connections i.e. parts of the network where audience members are more likely to remain trapped in their browsing behavior. Moreover, the study evaluates the extent to which audience behavior responds to that structure of news supply, using correlation techniques that control for dependencies typical in networks.


A more recent version of this study is going to be presented at MZES Conference, Mannheim on 23rd and 24th of October 2015.