The road to representivity

The rise of social media has profound consequences for those that research society to understand how it works. This paper is about trying to research social media in ways that both represent social media users, and also possibly wider society.

The rise of social media has profound consequences for those that research society to understand how it works. New bodies of information about society are now routinely produced through social media that are enormous, real-time, and rich new seams of evidence about social life: of people arguing, campaigning, talking about what they are doing, and offering what they think.

An important goal of social research is to study attitudes – to understand the fears, priorities, reactions and inclinations that people have and hold, whether to a person, a policy, company or idea. Attitudinal research is vital to inform decisions that affect and involve society: for agile, representative Government, for an effective and powerful civil society, and for good businesses that understand their consumers.

However, human attitudes are complex, difficult things to research. Principles and standards have built up over centuries to govern how to research them well, and how to avoid the many pitfalls that exist. One of the most important principles of social research, and especially when it tries to understand the attitudes of broader groups, is representivity. Research often depends on its ability to draw more general inferences about a wider population based on the data it uses, and the methods it employs. Representivity is not always important, of course, but it is vital when the research wishes to draw conclusions that are wider than simply those who have been the direct subjects of the research.

Researching society through social media must, in general, also reflect the principles of social science, including, in many contexts, the principle of representivity. However, representative research on social media is no easy task. New technologies are needed to understand very large, often complex social media datasets that are unfamiliar to social science, and do not easily fit within the conventional methods and frameworks that it uses.

This paper is about trying to research social media in ways that both represent social media users, and also possibly wider society. It is part of a wider effort between social researchers at Demos and Ipsos MORI, and technologists at the University of Sussex and CASM Consulting LLP, to build new and better ways of conducting social media research.

More insights about Technology & Telecoms

New Services