The key is understanding exactly how the topology and features of social networks can be used to expand the reach of positive interventionsand block the transmission of toxic influences.
Social Media Offers Help for the Hopeless
“Just can’t go on. Think I’m going crazy. Don’t see how I’ll still be here in a week or a month or a year.”
“Call the Lifeline. Don’t be shy. They can really help!”
These two lines from TrevorSpace, an online network that targets youth at high risk for suicide, capture the problem—and a potential remedy—at the heart of Vincent Silenzio’s research into using social media as a means of suicide prevention.
The first line is from a young person self-identified as LGB (lesbian, gay, or bisexual). Because of the social stigma they encounter, LGB young people 16 to 24 years old are three to four times more likely than other young people their age to attempt suicide or give serious thought to it, says Silenzio, associate professor of psychiatry, family medicine, and public health sciences.
Their isolation in society means they rely heavily on social media to establish networks of friends.
Silenzio sees in this a great potential to counteract the “toxic influences” that lead LGB young people to the depths of despair. He notes, “Their high rate of Internet use suggests that online social networks offer a novel opportunity to reach them”—to answer their despair with hope, for example, even if it is with a simple message of encouragement like the second line above.
Studies by Silenzio and other researchers show that large segments of “hidden” populations—including drug users and prostitutes— could be reached online by “peer-driven” messages. A public health message could be sent to a relatively small number of members recruited from that population, and the online network’s own connectivity—through “respondent-driven diffusion”—could take over to spread the word far and near with laser-like precision to the people it most needs to reach.
The key is understanding exactly how the “topology and features” of social networks can be used to expand the reach of positive interventions—and block the transmission of “toxic influences.”
That is where Big Data comes into play.
In one study in 2008, Silenzio and four other researchers used an automated data collection program—a “web crawl”—to examine people’s publicly accessible Myspace data to find those who openly identified themselves as LGB. The crawl was then extended to those individuals’ online friends—and the friends of those friends—until a network of 100,000 LGB individuals had been mapped.
A series of Monte Carlo simulations was run using computational methods to replicate what would happen if a real message were actually sent to members of the network. A variety of starting points was used. For example, what if various combinations of five, 10, or 15 randomly selected individuals were chosen to start the chain? What if they were given five or 10 coupons or alternative incentives to spread the message to other members of the network? The simulations showed that as many as 18,409 individuals could be reached.
One of the key findings: What matters most is not how many individuals the process starts with but the number of peers they recruit along the way. Increasing the number of coupons from five to 10, for example, caused a far more “dramatic increase” in the final sample size reached, compared to doubling or even tripling the number of initial participants.
Another of Silenzio’s goals is to develop social-enabled computer applications that could be put in the hands of teachers, clergy, and others who have close contact with LGB young people.
The applications would instruct them in the kinds of messages and support that can help point a distressed youth away from thoughts of suicide and would provide tools to disseminate these types of messages through social networks.
Ultimately, Silenzio hopes, his work in this area will become “superfluous” because society will have become more accepting of LGB young people and because stronger networks of support will be available to them.
At that point researchers will not have to devise ways to reach what is now a “hidden” population, with peer-driven messages launched online.