The Science Behind Using Online

Communities To Change Behavior

Sean D. Young
Saturday, September 28th, 2013
14 Comments

Editors note: Sean Young is a family medicine professor and director of innovation at the
center for behavioral and addiction medicine at UCLA. He writes about topics related to
psychology, technology, and public health/medicine at seanyoungphd.com.
Is it just me, or is it impossible to talk to technology entrepreneurs without mentioning user
engagement and behavior? Im a behavioral psychologist, so that might be why I keep having
conversations about engagement, but I dont think thats the only reason. I think its because
entrepreneurs have realized that behavior change and engagement is critical to technology
development (and to everything else in our lives). Whether were trying to get people to
download or keep playing our fantasy sports applications, convince ourselves to avoid that extra
scoop of ice cream, or get our neighbor to stop hitting her snooze button at 5:00, 5:10, and 5:20
a.m., we understand how difficult it can be to engage people and change behavior.
Before talking about how to do this, its important that were on the same page that being able to
change behavior is a powerful skill that could be used for good and/or evil. As entrepreneurs
seeking to improve the world, we first need to make sure that 1) the products we are creating will
benefit (rather than harm) society, and 2) the behavior(s) we want to change to be able to help us
achieve those objectives. Once weve gotten past those steps, how do we change behavior?
Fortunately, theres a science behind how to change behavior, and the answer to engagement and
behavior change lies in understanding peoples psychology. By addressing peoples
psychological needs and reasons for not changing behavior (including their social environments,
cultural values, and emotions), we can be more effective at behavior change. Once we
understand peoples psychologies, then technologies online communities in particular
become really useful as platforms to rapidly change behavior.

Changing behavior one social network at a time

Although social media and online communities might have been developed for people to connect
and share information, recent research shows that these technologies are really helpful in
changing behaviors. My colleagues and I in the medical school, for instance, created online
communities designed to improve health by getting people to do things, such as test for HIV, stop
using methamphetamines, and just de-stress and relax. We dont handpick people to join because
we think theyll love the technology; thats not how science works. We invite them because the
technology is relevant to them theyre engaging in drugs, sex and other behaviors that might
put themselves and others at risk. Its our job to create the communities in a way that engages
them enough to want to stay and participate. Yes, we do offer to pay them $30 to complete an
hour-long survey, but then they are free to collect their money and never talk to us again. But for
some reason, they stay in the group and decide to be actively engaged with strangers.
So how do we create online communities that keep people engaged and change their behaviors?
Our starting point is to understand and address their psychological needs.
In most of the studies, we first need to answer their questions about privacy issues (what is being
done with their data) and then build a trusting community. Unlike other communities, where
people typically join out of interest, we need to create their interest and keep them engaged
enough to want to stay and participate. To do this we have to slowly educate them about behavior
change so that they feel comfortable and not pushed into changing. Little things such as liking a
comment saying, Everyone having a good day? help them to feel safe and let them know they
can be themselves and have fun connecting with others, without being forced to talk about
health. After addressing these needs, we can then educate them and provide concrete, easy steps
for how they could change their behavior.
You might think that most people wouldnt actively participate in an online community unless
they wanted to be there in the first place and thought it was relevant to them. Well, results from
our three-month study released this September in Annals of Internal Medicine, found that over
80 percent of people who joined our groups were actively involved, and people in the
intervention study groups were more likely to change their behaviors compared to people who
were not. Interestingly, we keep finding that these research groups become actual online
communities; in our first study, the actual intervention ended over two years ago, but people
continue to actively use the community to reach out and support each other.
The Core Components
Throughout our research, we find that newly created online communities can change peoples
behaviors by addressing the following psychological needs:
The Need to Trust. Sharing our thoughts, experiences, and difficulties with others makes us feel
closer to others and increases our trust. When we trust people, were more open-minded, more
willing to learn, and more willing to change our behavior. In our studies, we found that sharing
personal information (even something as small as describing what you did today) can help
increase trust and change behavior.

The Need to Fit In. Most of us inherently strive to fit in. Social norms, or other peoples
attitudes and behaviors, heavily influence our own attitudes and behaviors. Each time a new
online community or group forms, it creates its own set of social norms and expectations for how
people should behave. Most people are willing to change their attitudes and/or behavior to fit
these group norms and fit in with the community.
The Need for Self-Worth. When people feel good about themselves, they are more open to
change and feel empowered to be able to change their behavior. When an online community is
designed to have people support and care for each other, they can help to increase self-esteem.
The Need to Be Rewarded for Good Behavior. Anyone who has trained a puppy knows that
you can get him to keep sitting as long as you keep the treats flowing to reward him, but if you
want to wean him off the treats and really train him then youll need to begin spacing out the
treats to make them less predictable. Well, people arent that different from animals in that way
and can be trained with reinforcements too. For example, liking peoples communications
when they immediately join a network, and then progressively spacing out the time that their
posts are liked (psychologists call this variable reinforcement) can be incorporated onto social
network platforms to encourage them to keep posting content. Eventually, these behaviors
become habits.
The Need to Feel Empowered. While increasing self-esteem makes people feel good about
themselves, increasing empowerment helps them know they have the ability to change. Creating
a sense of empowerment is one of the most powerful predictors of whether people will change
their behavior. Belonging to a network of people who are changing their own behaviors, support
our needs, and are confident in our changing our behavior empowers us and gives us the ability
to change our behavior.
Throughout my research career Ive found that understanding psychological needs is the core of
behavior change and engagement across all domains. People are complicated, so this list is by no
means comprehensive, but addressing these needs should get you a long way toward how you
can use technologies to change and sustain behavior.

Combo of social media, behavior psychology leads to HIV testing, better

health behaviors
Date: 09/04/2013
Contact: Enrique Rivero

Can social media be used to create sustainable changes in health

behavior?

A UCLA study published Sept. 3 in the peer-reviewed journal Annals of Internal

Medicine demonstrates that an approach that combines behavioral science with
social media and online communities can lead to improved health behaviors among
men at risk of HIV infection.

The evidence-based approach not only led to increased HIV testing and encouraged
significant behavioral change among high-risk groups but also proved to be one of
the best HIV-prevention and testing approaches on the Internet, according to the
study's lead investigator, Sean D. Young, an assistant professor of family medicine
and director of innovation at the Center for Behavior and Addiction Medicine at the
David Geffen School of Medicine at UCLA.

And it's not only applicable to HIV prevention efforts, he noted.

"We found similar effects for general health and well-being," said Young, who is also
a member of the UCLA AIDS Institute. "Because our approach combines behavioral
psychology with social technologies, these methods might be used to change
health behaviors across a variety of diseases."

In an earlier study, published in February and also led by Young, researchers found
that social media could be useful in HIV- and STD-prevention efforts by increasing
conversations about HIV prevention.

For the current study, the researchers recruited 112 men who have sex with men
through banner ads placed on social networking sites like Facebook, through a
Facebook fan page with study information, through banner ads and posts on
Craigslist, and from venues such as bars, schools, gyms and community
organizations in Los Angeles. Of the participants, 60 percent were African-American,
28 percent were Latino, 11 percent were white and 2 percent were Asian-American.

The men were randomly assigned to one of two Facebook discussion groups an
HIV intervention group or a general health group (with the latter serving as a
control in the study). Each participant was then randomly assigned to two "peer
leaders" within their group. The peer leaders communicated with participants by
sending messages, chats and wall posts.

In addition to general conversation, peer leaders for the HIV group discussed HIV
prevention and testing, while those in the control group communicated about the
importance of exercising, eating right and maintaining a low-stress lifestyle.

While the men were under no obligation to engage with the peer leaders or other
participants or to even remain members of their respective Facebook groups, the
authors found that the participants were highly engaged and maintained active
participation during the 12-week study.

Throughout the study, the men were able to request and receive home-based HIV
self-testing kits. At baseline and again after 12 weeks, participants completed a 92item survey that included questions about their Internet and social media use
(including whether they discussed health and sexual risk behaviors), their general
health behaviors (including exercise and nutrition), and their sex and sexual health
behaviors (including HIV testing and treatment).

Among other things, the researchers looked for evidence of behavioral change
such as reductions in the number of sexual partners and requests for homebased HIV test kits, along with follow-ups to obtain test results.

Among the findings:

95 percent of the intervention group participants voluntarily communicated

on Facebook, as did 73 percent of the controls.

44 percent (25 of 57) of the members of the intervention group requested

the testing kits, compared with 20 percent (11 of 55) of the controls.

Nine of the 57 intervention group participants took the test and mailed back
the test kits to receive their results, compared with two of the 55 control
group members, suggesting a greater likelihood that the approach can
successfully lead one to take an HIV test.

The intervention group members chatted and sent personal messages with
much higher frequency than did the control group members.

African-American and Latino men who have sex with men, who are at higher
risk for becoming infected with HIV compared with the rest of the population,
find social networks to be an acceptable platform for HIV prevention.

African-Americans and Latinos also find home-based tests to be an

acceptable HIV testing method.

In addition, the authors found that retention at follow-up was more than 93 percent,
in contrast to the high dropout rates from other Internet-based HIV-prevention
interventions.
"Internet HIV-prevention interventions and mobile health applications have had very
high dropout rates and problems getting people engaged, and this effect is even
more pronounced among high-risk groups such as minority populations and men
who have sex with men," Young said. "However, our approach appeared to
overcome these issues and led to changed behavior."

The researchers noted some limitations to the study, among them the fact that they
used only two Facebook communities per condition; these methods should be
tested with more people before implementing them, they said. In addition, no best
practices regarding the use of social networking for HIV communication have been
established.

The next step will be to assess how this method might generalize to other
populations, diseases, and prevention efforts, Young said.

"We have created a potential paradigm for health behavior change using new social
technologies," he said. "We are beginning to explore this approach in other areas."

Study co-authors were William G. Cumberland, Sung-Jae Lee, Devan Jaganath, Greg
Szekeres and Thomas Coates, all of UCLA.

Grants from the National Institute of Mental Health (K01 MH090884), the UCLA
Center for HIV Identification, Prevention and Treatment Services (CHIPTS) and the
UCLA AIDS Institute funded this study.

The UCLA Department of Family Medicine provides comprehensive primary

care to entire families from newborns to seniors. It provides low risk obstetrical
services and prenatal and inpatient care at UCLA Medical Center Santa Monica, and
outpatient care at the UCLA Family Health Center in Santa Monica and the MidValley Family Health Center, which is located in a Los Angeles County Health Center
in Van Nuys, Calif. The department is also a leader in family medicine education, for
both medical students and residents, and houses a significant research unit
focusing on health care disparities among immigrant families and minority
communities and other underserved populations in Los Angeles and California.

The Center for HIV Identification, Prevention, and Treatment Services

(CHIPTS) is a collaboration involving researchers from UCLA, Charles Drew
University of Medicine and Science, the Friends Research Institute, the RAND Corp.
and the broader Los Angeles community. It aims to enhance the collective
understanding of HIV research and to promote early detection, effective prevention
and treatment programs for HIV. Funded by the National Institute of Mental Health,
CHIPTS serves as a bridge among researchers, the government, service providers
and people with HIV in responding to changes in the HIV epidemic and in shaping
sound public policy. This is accomplished through a range of services, including
consultation on the development of new research projects and assistance with
obtaining funds for these initiatives. CHIPTS provides technical assistance in HIV
program development and evaluation and sponsors an annual conference for
researchers to present their work. In addition, the center hosts an annual policy
forum for researchers, government officials and the HIV community to discuss
emerging HIV policy issues and hosts a research colloquia series.

The UCLA AIDS Institute and Center for AIDS Research is a multidisciplinary
group of top-flight researchers united in the worldwide fight against HIV/AIDS, the
first cases of which were reported in 1981 by UCLA physicians. Institute members
include researchers in virology and immunology, genetics, cancer, neurology,
ophthalmology, epidemiology, social science, public health, nursing and disease
prevention. Their findings have led to advances in preventing and treating HIV as
well as other diseases such as hepatitis B and C, influenza, TB and cancer. To find
out about ways to support these efforts, please contact Laura Pescatore
(lpescatore@support.ucla.edu).

For more news, visit the UCLA Newsroom and follow us on Twitter.