Likes, followers, posts. These are the signs of popularity or success on personal social media pages.
When it comes to using social media to promote healthy behaviors, research has revealed that, in those spaces too, posting more often is generally associated with better outcomes. But the frequency of posts doesn’t tell the full story, according to new InCHIP Principal Investigator (PI) Sherry Pagoto. The content of posts may matter more.
“So many people are using online communities to manage their health, but so few people study what works best in them and how people can benefit the most from them,” said Pagoto, Director of the UConn Center for mHealth and Social Media. “We want to understand what types of posts by participants are most associated with healthy behavior changes.”
Pagoto, a clinical psychologist and Professor of Allied Health Sciences, moved to UConn and joined InCHIP this fall from the University of Massachusetts Medical School, bringing with her more than $6 million in federal grants, her Center, and its staff of 10. Pagoto’s collaborator, Molly Waring, an Assistant Professor of Allied Health Sciences and an InCHIP PI, also made the move.
The new InCHIP researchers are using a grant from the National Heart Lung and Blood Institute (NHLBI) to conduct a series of pilot studies examining engagement in online weight loss communities. Pagoto, the grant’s PI, will present their latest findings in early January at the Hawaii International Conference on System Sciences (HICSS), and the findings will be published in the conference proceedings.
The research team found that the three most common types of posts from a 12-week Facebook-delivered weight loss program were: (1) reporting a healthy choice such as “I just declined a doughnut. Go me!” (23%), (2) acknowledging another participant’s post such as “Great post!” (21%), and (3) sharing a challenge or slip-up such as “Ate way too much at the party!”(13%). The least common types of posts were: (1) negating such as “This recipe has broccoli. I hate broccoli” (1% percent), (2) irrelevant such as “Anyone see the game last night?” (2.5%), and (3) asking for help with a problem such as “I need ideas for low-cal snacks” (4%).
The types of posts most associated with weight loss were: reporting a healthy choice, asking for help with a problem, sharing a specific plan to eat better or exercise, and sharing progress on weigh-in days.
“We were intrigued that two types of posts that seemed similar – reporting a challenge and asking for help – actually were quite different in terms of predicting weight loss,” Pagoto said. “Reporting a challenge was one of the most common types of posts, but it wasn’t associated with weight loss. Asking for help was one of the least common types of posts but one of the most predictive of success. Asking for help is more proactive than just stating a problem and likely resulted in some ideas on how to move past the problem.”
Her team is testing ways to encourage participants to be more proactive in asking for help and to post content more likely to result in receiving personalized feedback and support from counselors and other participants.
After each study of online engagement, Pagoto’s team conducts focus groups with participants. Some have said they are not comfortable posting with strangers. Others are unsure of what to say.
“We are working on creating a supportive, confidential environment to make participants feel more at ease about participating. A weight loss group is also a new use of social media for a lot of people, so some might not be sure what is appropriate to post. We coach them on how to get the most out of these communities by telling them about what kind of posts are associated with greater success.”
Using popular social media platforms like Facebook to deliver behavioral interventions allows the researchers to reach the most people and to save time and money they otherwise would have to spend developing and marketing apps and websites. One potential drawback to social media delivery, however, is that people are used to sharing only good news on their personal social media pages, Pagoto said. This may account for why participants in the current study spent the most time reporting healthy choices they had made and less time asking for help.
“It’s a mental pivot for a lot of people, how to use one of these online communities differently than their regular social media pages,” Pagoto said. “When people do go out on a limb and share they have a problem, though, it’s actually very well-received. Other participants swoop in and offer advice, and they report that it made them feel better, not so alone.”
One of the research team’s next studies of engagement in online weight loss communities will test what happens when participants are allowed to invite family members and friends to join the community. The study will examine whether participants contribute more and ask for more help solving problems when they know people in the group and whether the number of people invited makes a difference.
While a number of her projects leverage existing social media platforms, Pagoto also is in the process of developing three mobile apps for weight loss.
“We’re moving toward integrating apps that help people with health behavior change, with social media communities that support them,” Pagoto said.