A network perspective on intermedia agenda-setting

Abstract

In Communication Theory, intermedia agenda-setting refers to the influence that different news sources may have on each other, and how this subsequently affects the breadth of information that is presented to the public. Several studies have attempted to quantify the impact of intermedia agenda-setting in specific countries or contexts, but a large-scale, data-driven investigation is still lacking. Here, we operationalise intermedia agenda-setting by putting forward a methodology to infer networks of influence between different news sources on a given topic, and apply it on a large dataset of news articles published by globally and locally prominent news organisations in 2016. We find influence to be significantly topic-dependent, with the same news sources acting as agenda-setters (i.e., central nodes) with respect to certain topics and as followers (i.e., peripheral nodes) with respect to others. At the same time, we find that the influence networks associated to most topics exhibit small world properties, which we find to play a significant role towards the overall diversity of sentiment expressed about the topic by the news sources in the network. In particular, we find clustering and density of influence networks to act as competing forces in this respect, with the former increasing and the latter reducing diversity.

Do you dare to travel to China? An examination of China's destination image amid the COVID-19

“…However, previous studies supporting such claims overlooked the influence of news media on social media. According to the Intermedia Agenda-Setting Theory, news media journalists, referred to as the “gatekeeper of information”, have significant influence on most topics and the sentiment expressed on other news sources ( Stern, Livan, & Smith, 2020 ). Moreover, communication studies empirically confirmed that online media outlets, like CNN.com , strongly affect other media, including social media ( Harder, Sevenans, & Aelst, 2017 ).…”

The impact of noise and topology on opinion dynamics in social networks

“…Diversity in media reporting is an important factor in promoting the reception of diverse ideas, especially for individuals with lower levels of education [ 36 ]. In a healthy society, it is important for the media to present a stable, balanced and diverse set of perspectives to the public [ 8 , 37 , 38 ].…”

“…We use the dataset collected by Stern et al [ 39 ], which contains a collection of texts produced by a list of 97 prominent online news sources between May and December 2016. Following the method outlined in detail by Stern et al [ 37 ], we use latent Dirichlet allocation (LDA) to identify 161 topics that are discussed throughout the eight-month period. For each of the topics, we then generate a sentiment index for each of the news sources that captures how the opinion of each of the news sources evolves over time for each of the topics.…”

“…In so-called intermedia influence networks, such as the ones we test here, previous work has demonstrated that a simple linear model can capture the relationship between the variance in opinions across news sources and macro-level network properties, e.g. the average clustering coefficient and density [ 37 ]. Having demonstrated that in synthetic data these network properties are encoded within the spectrum of the graph, we proceed by testing whether the expected opinion diversity is an indicator of realized intermedia opinion diversity.…”

“…where N k is the number of actors in the kth network and H t,i,k is the opinion of the ith actor in the kth network at time t. We train three linear regression models, each with weights learned using maximum-likelihood estimation. The first model is that proposed by Stern et al [37], and is of the form, log(y) = ω 0 + ω 1 L + ω 2 N + ω 3 C + ω 4 D, where L is the average shortest path length, N is network size, C is the average clustering coefficient, D is the network density. The second regression is of the form log(y) = ω 0 + ω 1 d, where d is the opinion diversity calculated using equation (3.9).…”