PERCEIVE project - Deliverable D5.4 "Report of the comparative analysis of the correlation between topics emergent from regional discourses on the one hand, and the awareness and perceptions of the EU (from Eurobarometer) on the other hand"

Abstract

The report at hand focuses on the statistical test of an eventual relationship between media representation of EU regional cohesion policy, among other explanatory variables, and individuals’ level of European identification as well as their different definitions of being European. At present time, it is largely acknowledged that individuals do not possess an innate sense of being European, rather, the meaning of such status is socially constructed. Accordingly, extant research has explored the role of the media in shaping the opinions of the general public. However, only recently has research started focusing on the potential role of cohesion policy in shaping EU identity and many aspects of this phenomenon are still unexplored. We claim that by extending knowledge in this still unfolding area our work contributes to the wider debate on European identity in several ways: a) by performing a media analysis in seven different countries we offer one of the first international evidences as most of the media analyses are conducted in individual national contexts; b) our study is based on a large ad-hoc designed survey which allows us to capture so far largely overlooked aspects of EU identity such as the multiplicity and synchronicity of levels – i.e. individual, regional and national – in a way that Eurobarometer-based research could not do so far; c) our survey also allowed us to explore in unprecedented depth the factors associated with different definitions of being European; d) we analyse media in a bottom-up way, that is, without using pre-coded frames of valence characterising most of extant research; e) along with standard interpretive techniques, we use formal methods for representing national media spaces such as topic modelling and sentiment analysis. We believe that this mixed method approach makes our study more replicable than purely qualitative ones, still, nuances and complexities of extracting meaning from text are better preserved than purely quantitative studies;