Abstract

GODINHO, Luísa. Big Data and politics: determining the predictive potential of the world wide web in the European elections of 2014. OBS* [online]. 2018, vol.12, n.2, pp.256-285. ISSN 1646-5954.

Inversely to the dominant trend in predictive electoral studies in a digital environment, focused on the microblogue Twitter as a universe of analysis, this article develops an integrated ecological approach to the positioning of Portuguese public opinion during the electoral campaign for the European elections of 2014. The predictive potential of three categories is tested - party / coalition; head-of-list and list - as well as is the representativeness of the world wide web, considered a Big Data scientifically valid for Social Sciences research. The first category revealed a higher relative empirical value and the suitability of social networks for the anticipation of events is partially reinforced. However, there is a need for further studies, especially at the national level, where research in this area is practically non-existent.

Keywords : Big Data; Internet; elections; public opinion; positioning; electoral forecasting.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )