Rhetorical manifestation of institutional transformation

The present study proposes an analysis process to compare and contrast different approaches to content analysis. Moving from previous findings (Righettini and Sbalchiero, ICPP—international conference on public policy, 2015), related to consumer protection in the annual speeches of Italian Presidents of AGCOM, delivered between 2000 and 2015, statistical analyses of textual data are applied on the same set of texts in order to compare and contrast results and evaluate the opportunity of integrating different approaches to enrich the results. This review of results resorts to topic based methods for classification of context units (Reinert, Les Cah l’Anal Donnees 8(2):187–198, 1983), text clustering and lexical correspondence analysis (Lebart et al., Exploring textual data, 1998) in a general framework of content analysis and “lexical worlds” exploration (Reinert, Lang Soc 66:5–39, 1993), i.e., the identification of main topics and words used by AGCOM Presidents to talk about consumer protection. Results confirm the strengths and opportunities of topics detection approach and shed light on how quantitative methods might become useful to political scientists when available policy documents increase in number and size. One methodological innovation of this article is that it supplements the use of word categories in traditional content analysis with an automated topics analysis which exceeds the problems of reliability, replicability, and inferential circularity.