Does ‘bigger’ mean ‘better’? Pitfalls and shortcuts associated with big data for social research

‘Big data is here to stay.’ This key statement has a double value: is an assumption as well as the reason why a theoretical reflection is needed. Furthermore, Big data is something that is gaining visibility and success in social sciences even, overcoming the division between humanities and computer sciences. In this contribution some considerations on the presence and the certain persistence of Big data as a socio-technical assemblage will be outlined. Therefore, the intriguing opportunities for social research linked to such interaction between practices and technological development will be developed. However, despite a promissory rhetoric, fostered by several scholars since the birth of Big data as a labelled concept, some risks are just around the corner. The claims for the methodological power of bigger and bigger datasets, as well as increasing speed in analysis and data collection, are creating a real hype in social research. Peculiar attention is needed in order to avoid some pitfalls. These risks will be analysed for what concerns the validity of the research results ‘obtained through Big data. After a pars distruens, this contribution will conclude with a pars construens; assuming the previous critiques, a mixed methods research design approach will be described as a general proposal with the objective of stimulating a debate on the integration of Big data in complex research projecting.