Moving up. Applying aggregate level time series analysis in the study of media coverage

Abstract

In this article the advantages of aggregate level time series analysis for the study of media coverage are discussed. This type of analysis offers the opportunity to answer questions relating to causes and effects of media attention for issues and all kind of other content characteristics. Data that ask for a time series approach have become widely available during the past years, due to the rise of digital archives and social media such as Twitter and Facebook. This type of analysis allows for answering a set of interesting research questions and strong inferences about causal processes. Common challenges in time series analysis, relating to stationarity, accounting for a series’ past and autoregressive conditional heteroscedasticity are discussed. Two useful approaches, ARIMA and VAR, are introduced stepwise. An empirical example, dealing with intermedia agenda-setting between different newspapers in the Netherlands, demonstrates how both techniques can be applied and how they provide insightful answers to interesting research problems.