An uproar over the release in May of data on nearly 70,000 users of an online dating site without their permission serves as an important reminder to researchers that they need to ensure adequate ethics approval before delving into the world of social media, experts say.
Professor Deborah Lupton, a sociology and media studies researcher at the University of Canberra who focuses on social and cultural aspects of digital health, said ethics committees worldwide were struggling to keep up with ethical guidelines as researchers turned increasingly to digital and social media to recruit and engage with their cohorts.
“There are so many opportunities there you would be silly at least not to think about them, but it’s up to researchers when they are planning their research to be aware of all these issues they may be criticised for, and to try and do things as rigorously as possible.
“It’s very contextual, you can’t just have broad guidelines about ethical issues. You need to think about the topic, target audience and the social medium itself, and all of this has to be considered for each project,” she said.
The OkCupid scandal
The ethics around the use of social media for research was thrust into the spotlight recently when a group of Danish computer scientists published data collected from November 2014 to March 2015 on nearly 70,000 users internationally of the OkCupid dating site, including their usernames, gender, personality traits and sexual preferences.
The researchers posted the data on the Open Science Framework, arguing that it was publically available. However, they did not attempt to anonymise the data, or to contact OkCupid or the users to ask for permission to use it.
They were quickly condemned by the broader scientific community as having violated research ethics, and their paper was removed. However, it remains unclear whether the exercise breached any existing privacy laws.
In the absence of clear ethical guidelines governing the use of the internet for research, “public health researchers need to be highly cognisant of the ethics of what they are doing and that they might be very different to what they are used to”, Professor Lupton said.
Social media ethics – a personal perspective
Professor Kate Steinbeck, Medical Foundation Chair in Adolescent Medicine at the University of Sydney, and her team have been using social media and SMS messaging for several years in their ARCHER study, a longitudinal study on the effects of puberty hormones on adolescent behaviour, mood and wellbeing.
Professor Steinbeck said the best recruitment method for younger adolescents aged 10 to 12 had been through schools and emails to parents, though Facebook had been a useful tool for keeping families engaged, and text messaging had kept the participation rate at about 90% for more than 3 years.
“What we found with Facebook was that people really liked the pictures. We learned that if we wanted to get people’s attention, it wasn’t about telling them about the research, it was linking it to a picture, which they were more likely to ‘like’ and send to others,” she said.
“We also found that email was better than telephone, but people really like face-to-face contact. However we don’t really have the data to understand the contact dose of face-to-face vs electronic communication.”
She cautioned that social media was by its nature interactive, and it was important to understand boundaries around researchers’ capacity to respond to comments or text messages, especially in cohorts of vulnerable individuals.
“There can be very significant privacy issues if you start putting two-way conversations on media platforms, and you have no control over where the information goes.
“Also, you really need to understand when you’re going to respond to a text message – what if you received an SMS saying a teenager was thinking of suicide and it was downloaded immediately into a study database?” she said.
Recruiting and engaging with cohort participants is not the only area of cohort research that is seeing new developments. The latest issue of Public Health Research & Practice explores how researchers are also engaging with cohort data in new and innovative ways.
Tips for researchers
Dr Becky Freeman of the School of Public Health at the University of Sydney, whose research interest is how social media is being used by public health, said it was now becoming impossible to obtain an accurate representation unless researchers considered social media.
However, new technology was seldom used in a strategic way by public health researchers, she said.
She and others suggested the following tips for public health researchers to get the most out of digital media:
- If you don’t have a sophisticated understanding of social media, collaborate with or engage someone who does.
- Make sure you get the ethics right. The Association of Internet Researchers has published ethical guidelines governing research with digital media and apps. View them here.
- Consider privacy and confidentiality issues, and how young people, in particular, view privacy on the internet.
- Consider the possibility of two-way interaction, especially with vulnerable cohorts. Do you have capacity? At what stage does research become an intervention?
- If you use digital media, clearly outline the process in your paper so your methods can be evaluated. Consider writing a methodological paper afterwards to add to the evidence base.
- Do not assume that social media is the only way of engaging. Face-to-face interviews may elicit more interesting details.
- Remember that social media platforms are fluid: the most popular forum may change during the course of the study period.
- Read more about the internet and research in the open access Journal of Medical Internet Research.
