Social media data use needs tighter research controls, experts say

Experts have called for tighter control of the practice, with fresh guidelines needed to ensure personal data is being used appropriately.

Researchers at the University of Edinburgh say ethics frameworks around consent, privacy and ownership for such studies are not keeping pace with technological developments.

The plea comes as more and more social media data is used by researchers to reveal valuable insights into our behaviours, feelings and opinions.

Advances in tools to draw patterns from large datasets have opened the door to research projects that mine this deep seam of information.

Such techniques are, for example, being used to probe whether people's social media updates can predict the onset of mental health problems.

The authors found that, out of 13 sets of ethics guidelines developed or endorsed by Research Councils UK, only four explicitly mentioned the use of social media data in research.

They also analysed 156 published health studies that involved social media data and discovered that less than a third reported having made any ethical consideration of the use of personal information.

Only two of the nine studies from UK institutions made reference to RCUK recommended guidelines.

The research is published in the journal Research Ethics.

Dr Claudia Pagliari, of the University of Edinburgh's Usher Institute, said: "Our study highlights a significant gap in UK guidance on mining social media data for research purposes. Funding bodies, learned societies, research organisations and journals -- in addition to the researchers themselves -- all have a role to play in ensuring such research is carried out to the highest ethical standards."

"Ethics is about more than privacy in this context. Researchers may be using information that has been willingly shared in the public domain but this doesn't give them carte blanche to do as they please. Asking permission to use people's social media postings is courteous, although this may be impossible in very large studies. Treating personal information with confidentiality and respect, and avoiding its misuse for unethical purposes, are essential.

"Having good interdisciplinary guidelines and clear expectations for how these should be applied will help to improve practices."