Last Wednesday, CLT ran a workshop on Exploring Social Media as data sources for research as part of our NetworkEd series.

There was an excellent turnout of around 35 academics and PhD students from across the LSE’s departments attending, which shows that there is real interest in developing effective research methods to analyse the wealth of data social media can provide.

Some of the tools we explored can be found here. We got the ball rolling by discussing the advantages and disadvantages of using social media data for research, which Jane presented below:

advantages and disadvantages of social media for research data

Common advantages we identified included the convenience and abundance of data available to researchers. However, the disadvantages of using this data include issues with the accuracy and veracity of information; issues which must be considered when conducting any research, but which are conflated by the amount of data, and the speed at which it is generated. We also discussed several aspects of using social media data for research, including:

Ethical challenges

  • Just because data is accessible, does it make it ethical?
  • What is the human subject in an online context
  • Who are you researching, and do you have their consent? Getting informed consent from anonymous, pseudonymous and group profiles can be a huge challenge.
  • Are participants aware of the aims of the research being conducted and do your outputs align with their perceptions?
  • Is there a potential to victimise people and groups on social media through your research?

Legal issues

  • Ambiguity surrounding the legal frameworks on getting information from individual and group profiles; could there be potential for libel and misrepresentation due to the blurring of public and private roles on social media platforms?

Methodological issues

  • Is social media an appropriate data source for your research question? For example, lack of internet in certain areas, or social media access or state clampdowns on seditious activity online could make fair comparisons between situations in different countries difficult or even impossible.
  • Analysing vast amounts of data effectively can be time-consuming and sourcing appropriate data requires vigilance.
  • Could observing individuals and groups on social media affect their behaviour?

We also heard a couple of great case studies on how data can effectively be mined from Twitter, Weibo and other platforms, and how accessible, powerful analytical tools can provide fantastic insights, particularly when studying network interactions.

In summary, while ethical, legal and methodological issues exist around using social media data for research, researchers, academics and academic support staff at the LSE are making progress in developing tools and frameworks for effective research using social media, and the ever-changing nature of social media means that this process is likely to be an ongoing and iterative process.

Indeed, our own activities in this area have only just begun, and we would like your feedback on this event, as well suggestions for tools or events you feel would be useful towards your research, which you can leave as comments below or via Twitter (@LSECLT or #LSENetEd).