Love Data Week: Is using social media data in research ethical?

Experts met virtually this week to discuss the ethical use of social media data in research in a panel hosted by the College of Arts and Sciences’ Center for Ethics and Human Values at The Ohio State University.

The conversation, sponsored by the center’s Conversations About Research Ethics program (CARE), was one of many held at the university during Love Data Week, an international celebration of research data.  

Moderator Dana Howard, assistant professor in the College of Medicine Division of Bioethics, kicked off the panel by sharing the theme for this year’s Love Data Week: Whose data is it anyway?

This question was at the heart of the conversation between the three panelists, Casey Fiesler, associate professor of information science at the University of Colorado Boulder; Erin Moore, assistant professor of anthropology at Ohio State; and Annie Specht, associate professor of agricultural communication at Ohio State.

Fiesler first walked the group through her work with data from the social platform X, formerly known as Twitter. Citing the work of another researcher, Fiesler compared X to a fruit fly.

“People don’t study the fruit fly because it’s the best thing to study. It’s not because fruit flies are the most like humans,” she said. “They are easy to study for various reasons: short lifespan, easy to keep alive in the lab. [X] is similarly not the most representative of the population but it’s very easy to get [X] data.”

Specht furthered this point later in the conversation.

In addition to being easily accessed, social media, she said, provides “a real-time snapshot of folks’ opinions, beliefs and attitudes about what they’re seeing.”

It makes sense, then, that researchers would seek out social media data for their research. All three panelists agreed that using the data was acceptable, as long as it was done ethically.

How best to do that? Consider the community you’re entering, Fiesler said.

“What are the norms and expectations of the people who are represented by the data? How could individuals or communities be harmed by using it … or disseminating it beyond its original context?” she said. “Who benefits from this use and are the people benefiting the same ones who are facing potential harms? … What could you do to mitigate the potential for harm or increase the likelihood for the benefits of this work?”

With those questions in mind, researchers need to think long and hard about how they’d answer them, she said.

“Can you answer those questions well enough to suggest that you’re the right person to be conducting that research?”

Moore spoke to the concerns that social media data isn’t reliable.

“Irving Goffman, this old-school sociologist, had this idea that people are performing all the time, that there’s this performance of everyday life,” she said, which suggests that people aren’t being honest in their posting.

Moore’s recent work focused on TikTok videos made by the parents of children with type 1 diabetes. The videos show the psychological toll managing their child’s illness takes. Insulin is administered via injection, so children are frequently in pain and upset.  

“To say something is performed doesn’t make it any less real, right?” she asked. “… Of course, these were produced videos. The parents had to set the phone a certain way to show bruises on kids’ arms. There was a lot that went into making that production there. The production can still be analyzed. It’s like its own artifact.”

Howard asked if there was an argument to be made that users should know their data is publicly available because of the terms of service agreements they sign. Fiesler vehemently disagreed.

“Terms of service are completely useless for making any kind of ethical decision,” she said. “They don’t have anything to do with ethics and they rarely mention research at all.”

In closing, Moore argued that the fluidity of social media means that ethical researchers must be proactive and adaptable.

“I think that involves thinking through the ethics at every step,” she said. “The ethics don’t get decided on the front end and then you go forward. You have to constantly think about it at every step.”