Keywords: Social media, big data, sentiment analysis, and emerging technologies

Internet impact in civic engagement has been widely studied by the literature (see Boulianne, 2009). However, most of the efforts to determine the effects of social media have been concentrated in the behavioural aspects of the engagement (see Boulianne, 2015). Although many studies incorporate political attitudes on their models (e.g. Carlisle & Patton, 2013; Gil de Zúñiga et al., 2017, 2012, 2014; Prior, 2005; Valenzuela et al., 2009; Yang & DeHart, 2016), we identified, at least, two problems: a) Political attitudes – specifically internal efficacy and external efficacy – are not methodologically treated as literature suggest; and b) Political attitudes are usually incorporated into models to control, but are rarely treated as dependent variables.

[Objective] This work seeks to determine the impact that different uses of social media would have on the equitable development of certain political attitudes usually associated with disaffection: Internal Efficacy, External Efficacy and Political Interest (Maldonado Hernández, 2013; Torcal and Montero, 2006). In this way, we seek to contrast the following hypothesis:
H1: Civic use of social media is more connected with internal efficacy than recreational use of social media.
H2: Civic use of social media is more connected with external efficacy than recreational use of social media.
H3: Civic use of social media is more connected with political interest than recreational use of social media.

[Methods] To contrast our hypotheses, we use a survey designed by us and implemented by GFK-Adimark. The survey – composed by 1650 face-to-face cases – includes the 10 most populated cities of the Chilean Region of Valparaíso. With this data, and following the literature (Bollen, Rabe-Kesketh, & Skrondal, 2008), we estimated different Structural Equation Models (SEMs). This method allows us to simultaneously build our factors and estimate the relationship between our variables. The results of our analysis suggest that the civic use of social media would be more related than recreational use to the development of internal and external efficacy. In other words, the use of social media with an informational or political purpose would increase the sensation of self-competence and the sense of system responsiveness. Based on this, we understand that the different uses of these new platforms could not only generate knowledge or participation gaps, but also attitudinal gaps between citizens.

[Added value] We consider that the object of this investigation and its methodology represent an important contribution to determine the interaction between the social media and the attitudinal elements of the civic involvement. Although there have been important efforts to analyze the true democratizing power of the Internet and social media, we understand that our work presents some differentiating characteristics that we consider essential for the good understanding of the democratizing effects of new technologies.