Geo-social media and socio-spatial stratification: new methods for a conjoint analysis

The role of space in sociological theory has never been preeminent or in any case well defined. Despite the works of authors such as Giddens (1979), Harvey (1978), Gieryn (2000) and Jacobs (1961) highlighted the centrality of the spatial dimension in the study of social phenomena, this relevance is not yet fully recognized except for the field of urban sociology, where the focus on the space and territorial dimensions are fundamental keys to understand and explore social inequalities (Tammaru et al., 2016; Vicari Haddock, 2013). Recently, the interest in the spatial dimension seems to have found new life thanks to the great diffusion of big data in the social research. In the data revolution era (Kitchin, 2014) new data and new sources allow researchers to find new ways to study society and its dynamics. Among these types of data, geo-located data enable better ways of producing social knowledge (Halford, 2013). The availability of "voluntary" (Flanagin, Metzger, 2008) and "derived from social media" (Campagna et al., 2015) geographic information put the spatial dimension – initially ignored in social media analysis – at the center of the interest in digital and web studies. In the last few years, a new approach which aim to analyze two worlds that were previously considered irreconcilable, the online and the offline world, has been developed. To date it is increasingly recognized that the virtual and material separation is entirely artificial, as pointed out in spatial media theory (Leszczynski, 2014). In the growing body of research on geo-social media it is possible to distinguish at least five lines of research. In one of these we can find quantitative works which focus on explaining the variability of the phenomena investigated through the analysis of geo-social media, by using statistical models where socio-economic variables – gathered from ecological units – are treated as independent variables. This field of research investigates the different ways by which the spatial dimension is related to events of online world and in particular of social platforms. Starting from this consideration, the work aims to illustrate the potential of some innovative methods for the analysis of georeferenced data to integrate the study of socio-spatial stratification with the social media analysis, through the adoption of a large geolocated twitter dataset collected from 2017 to 2020 and the use of spatial analysis techniques performed with R and Qgis software.