تجزیه و تحلیل محتوا از پوشش روزنامه از احیای گرگ در فرانسه با استفاده از مدل سازی موضوع ساختاری

Populations of large carnivores are recovering in Europe and incur increasing conflict interactions with human activities. According to the agenda-setting theory, the media dissemination of information on these interactions is likely to contribute to shaping public perceptions of large carnivores. We conducted a content analysis of printed media coverage of wolf recovery in France over the period 1993–2014, ever since its natural return to southeast France. To do so, we used a recently developed statistical method – structural topic modeling – that allows to generate topics from large amount of texts and formulate new or assess existing hypotheses. This method formally includes covariates to explain variation in topic prevalence and content in a way that is similar to standard regression analyses. We contrasted content variation between articles in a regional (Nice-Matin; N = 742) and a national (Le Monde; N = 148) newspaper and analyzed time trends in topic prevalence. The most represented topics were mainly related to the management issues regarding wolf recovery. We found that Le Monde represented management issues in a generic manner associated with a perspective centred on carnivore species. In contrast, articles in Nice-Matin were about factual issues and associated with a human-centred viewpoint. This contrasted framing emphasizes the gap in representations of wolf management between citizens who directly interact with the wolf and favor detailed information content, centred on human views, and citizens who do not interact or only indirectly with the wolf who will focus on less detailed news, with a more ecological approach. We suggest that increased communication between local and national stakeholders and institutions could provide the context for a more balanced media content of interactions between carnivore species and human activities. This combination could attenuate the gap between regional and national representations.