The social networking site LinkedIn reported a fall in the value of its shares last week and, although predictions of failure surface every couple of months about the big-hitters, Facebook and Twitter, social media websites don’t seem to be losing any of their popularity. Which is good for us researchers, as such websites present endless opportunities for recruitment, communication, observation and ‘scraping’.
But what about the millions of individuals who use these sites –what do they think about their posts, ‘likes’ and statuses being used for research purposes? Over the last year we, at NatCen Social Research, have sought to answer this question by undertaking research with social media users in London. This week we published a report outlining our findings and setting out practical recommendations for any researchers wanting to use social media websites. You can read the report in full, but here’s a taster of what we found:
Types of user
Wide variation exists in how individuals use social media websites. There are ‘observers’, for example, who only view the posts and activity of others; others who create and disseminate their own content (‘creators’) and those who prefer to ‘re-post’ existing material (‘sharers’). One individual can also be a mixture of all three, depending on which platform they are using. Researchers should consider if they want to represent views from all these types, and how they could do so.
As well as differences in how they use sites, social media users also differ in how familiar and comfortable they are with the privacy and security settings that are provided by the websites. Researchers may want to help any potential participants understand how public or private any of their discussions are, and how traceable they might be if mentioned in a research report.
Views on research using social media
Our participants held different views about the idea of researchers using social media:
- Scepticism was expressed by those who thought the traditional research methods (for example face-to-face interviewing) were seen as effective and so, ‘why fix what ain’t broke’?
- Others accepted that researchers would want to use social media and that it could overcome some of the barriers that traditional methods have –like interviewer bias, whereby the interviewer unknowingly has an influence on what respondent talks about.
- A final group assumed that such research was probably happening already and were fairly ambivalent about it as a result.
There was also concern about how social media users would be able to tell that a researcher was ‘genuine’ if they approached you online. To build trust, it was suggested that researchersapproach potential participants by using the private message facilities provided by the websites, rather than finding and using an email address or phone number from a profile, which would seem as if the researcher had already been ‘snooping’ around.
Informed consent and anonymity
Participants were particularly concerned by the idea of their photos, Twitter handles (or other usernames) or particularly private or sensitive statuses being published in research reports. Ideally they wanted researchers to ask for their permission to use these, although there was recognition of the practical problems this could pose if the researcher had ‘scraped’ data from hundreds, even thousands, of profiles. On the other hand, there was a view that if you had posted something on social media it was ‘fair game’ if it was taken and used, because only you can control how privately you post anything.
Continuing the discussion
We would be interested to hear readers’ views about our research and also about how feasible they think our recommendations are. If you would like to learn more about social media research, NatCen co-coordinates the New Social Media, New Social Science (NSMNSS) network, which will continue to raise questions and offer solutions in this area.