620392 research-article2015
HPPXXX10.1177/1524839915620392Health Promotion PracticeEdgerton et al. / New Strategies to Assess Online Health Communication
Social Media & Community-Engaged Practices
Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention Elizabeth Edgerton, MD, MPH1 Erin Reiney, MPH, CHES1 Siobhan Mueller2 Barry Reicherter2 Katherine Curtis2 Stephanie Waties2 Susan P. Limber, PhD3
Every day in classrooms, playgrounds and school hallways, through text messages and mobile technology apps, children are bullied by other children. Conversations about this bullying—what it is, who is involved, and how to stop it—are taking place online. To fill a need for relevant, research-based materials on bullying, the U.S. Department of Health and Human Services’ Health Resources and Services Administration worked with Widmeyer Communications to investigate the scope of media conversations about bullying and discover new strategies for promoting appropriate public health messages about bullying to intended audiences. Key components of the methodology included: analyzing common search terms and aligning social media content with terms used in searches rather than technical language; identifying influencers in social media spheres, cultivating relationships with them, and sharing their positive, relevant content; examining which digital formats are most popular for sharing and creating content across platforms; tracking and reporting on a wide variety of metrics (such as click-through and engagement rates and reach, resonance, relevance, and Klout scores) to understand conversations around bullying; and looking at online conversations and engaging participants using applicable resources and calls to action. A key finding included a significant gap
Health Promotion Practice May 2016 Vol. 17, No. (3) 448–456 DOI: 10.1177/1524839915620392 © 2016 Society for Public Health Education
between search terms and online content and has led to recommendations and comprehensive ideas for improving the reach and resonance of StopBullying.gov content and communications. Keywords: social marketing; social media; health communication; metrics and evaluation; bullying prevention; online influencer identification; keyword research; conversation mapping
B
ullying is an all too common experience in the United States, with more than one in five youth between the ages of 12 and 18 years targeted at school (Robers, Zhang, Morgan, & Musu-Gillette, 2015). It is defined as unwanted aggressive behavior among school-aged children that “involves an observed or perceived power imbalance and is repeated multiple times or is highly likely to be repeated” (Gladden, VivoloKantor, Hamburger, & Lumpkin, 2014, p. 7). Longitudinal research has found that chronic bullying impairs victims’ health, but identification and early intervention 1
Health & Human Services, Health Resources and Services Administration, Rockville, MD, USA 2 Widmeyer Communications, Washington, DC, USA 3 Clemson University, Clemson, SC, USA Authors’ Note: Address correspondence to Siobhan Mueller, Widmeyer Communications, 1129 20th Street NW, Suite 200, Washington, DC, USA; e-mail: [email protected].
448
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
by clinicians can stem negative effects (Bogart et al., 2014; Gini & Pozzoli, 2013). Because of the high prevalence of bullying and the serious short- and long-term effects identified in children who are bullied (Ttofi, Farrington, Lösel, & Loeber, 2011a), children who bully others (Ttofi, Farrington, Lösel, & Loeber, 2011b), and even those who witness the behavior (Espelage & Swearer, 2010), bullying is recognized as a public health issue. In 2001, the Heath Resources and Services Administration (HRSA) joined efforts to address bullying in schools and communities by launching a social marketing and communications campaign to promote prevention and intervention strategies. HRSA selected Widmeyer Communications to facilitate the planning, design, production, and dissemination of the initial campaign materials and continues to develop content and support communication strategies. In 2011, HRSA joined forces with other government agencies to coordinate efforts and establish the Federal Partners in Bullying Prevention, which includes the departments of Education, Defense, and Justice, among others. The Federal Partners launched StopBullying.gov, a website with research, best practices, and training materials for bullying prevention and response. Stakeholders include, but are not limited to: parents, educators, health and safety professionals, law enforcement officers, and child-care providers. Social media messaging was also introduced through a Facebook page and Twitter profile to complement the main portal of web-based information at StopBullying.gov by bringing attention to the site and promoting its content. The goals of these digital assets were to expand the number of community stakeholders being reached, and to further connect the information and best practices available on the website with the target audiences. Traditional metrics such as number of site visits, Twitter followers, and Facebook likes were used to track progress over time. Data indicated that StopBullying.gov platforms experienced consistent but moderate growth between 2011 and 2013; however these indicators were limited because they only showed numbers of people who used online resources. They did not accurately indicate whether people who searched for help addressing bullying found the information they needed from the site or its social media pages.
Literature Review >> In the past, using social media to promote public health campaigns or other socially beneficial concepts has been primarily restricted to pushing out untailored, one-way messages (Neiger, Thackeray, Burton, GiraudCarrier, & Fagen, 2013). Few organizations have mastered
the art of meaningful exchanges with constituents or consumers, and as a result, they fail to take advantage of the interactive nature of social media. For many social media managers with government agencies, the slow pace of social media adaptation does not stem from a lack of understanding or desire, but from restrictive policies imposed on content and postings. Media richness theory proposes that media differ in their abilities to facilitate changes in understanding among communicators (Daft & Lengel, 1986; Kahai & Cooper, 2003). As Ledford (2012) notes, social media channels rank high on both richness (quick response, ability to communicate multiple cues, ability for natural language, and personal focus) and content control. Because of content control challenges, the nature of social media by public agencies tends to focus on developing and sharing their own content, responding to comments or questions, sharing content from nongovernmental sources, and retweeting content or links from private citizens. Comments and feedback are moderated and shared content is controlled, lacking true engagement (Thackeray, Neiger, Burton, & Thackeray, 2013). The potential of social media as a meaningful influencer of public health promotion is evident in the numbers: Researchers estimate that by 2017, the number of individuals and corporations who have social networking accounts will reach more than 4.8 billion. In particular, Twitter is becoming ever more popular. A total of 18% of Internet users have a Twitter account . . . [It] has been used to disseminate information about diabetes, breast cancer, to communicate during a disaster, and to understand health-related trends and issues such as influenza, tobacco, prescription drug misuse, and suicide. (Thackeray et al., 2013)
Still, social platforms are not yet being used to their full advantage by health departments (Thackeray et al., 2013) and other public agencies. The most successful social marketing campaigns leverage social media to its full capacity. A good example is The Heart Truth, the first national awareness campaign designed to educate women about their risks of heart disease, the leading cause of death among U.S. women. Sponsored by the Health and Human Services’ National Heart, Lung, and Blood Institute, the campaign harnessed social media marketing and digital strategies in the campaign’s design and in the highly successful Red Dress brand that continues to be used 5 years later by the National Wear Red Day and the Red Dress Collection (Long et al., 2009). According to Long et al. (2009), The Heart
Edgerton et al. / New Strategies to Assess Online Health Communication 449
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Truth campaign leveraged a range of social platforms and applications, including blogs, Facebook, YouTube, Wikipedia, and Flickr, as well as online public service announcements and pay-per-click ads. Targeted outreach was also conducted to influential bloggers who write about women, health, motherhood, fashion, and entertainment. The Heart Truth campaign is credited with having increased women’s awareness about heart disease and its prevention and contributed to a reduction in rates between 2002 and 2008 (Long et al., 2009). Lilley (2007) describes the potential of applying commercial marketing techniques to the social sector in this way: Just as a large chain store tracks consumers’ purchases and predicts future needs as a result, public health departments could target individuals interested in smoking cessation or weight loss with appropriate messages and resources. A study by Harris, Mueller, and Snider (2013), which examined the use of social media by local health departments in promoting health messages, concluded the practice is not widespread, but it is emerging especially in large metropolitan areas. A cross-sectional study of state public health departments by Thackeray, Neiger, Smith, and Van Wagenen (2012) confirmed that adoption of social media by these agencies is still in the early stage and primarily used as a digital version of traditional oneway dissemination. Most health departments are not optimizing the potential of social media platforms for interactivity and engagement that could result in improved health outcomes (Thackeray et al., 2012). Developing customized digital strategies for using, evaluating, and revising social media activities that are engaging and interactive is still very unique. Moreover, to our knowledge, no studies of social media strategies for bullying prevention have been published to date. Recognizing these gaps in the field, the purpose of this study was several fold: (a) investigate the scope of social media conversations around a critical public health issue—bullying, including reach resonance, and relevance; (b) discover new strategies to leverage social media more effectively for the promotion of research-based public health messages about bullying and best practices in bullying prevention; and (c) determine the extent to which this content is likely to reach new key target audiences.
Method >> In 2013, HRSA collaborated with Widmeyer Communications to conduct this research which included the collection of data from 2011-2012. The research approach was divided into two broad categories:
••
••
Defining the conversation by identifying relevant audiences and topics, creating a seed term list with popular search terms and keywords, and then creating a key search term list based on examining Google search data. For the purposes of this research effort, conversation is defined as any public social or digital media communication about bullying that falls within the keyword query parameters. The conversation was analyzed on Twitter, Facebook, blogs, discussion forums, news sites, and YouTube. The conversation is composed of individual tweets, blogs, articles, and comments shared on the aforementioned platforms. Comments, which can be left on news articles and blogs, are defined as their own conversation category in the analysis. Research and analysis consisting of conversation targeting, monitoring, and data collection; analysis; and message creation. The tool used to conduct the conversation analysis, Crimson Hexagon, collects data through three methods: streaming, pinging, and crawling. Streaming data is the most direct method and provides real-time, unrestricted access to the Twitter content stream. Pinging content collection, or API (application programming interface) call, is used on Facebook, Instagram, and Google+. Pinging also provides direct access to platform data, but the source may place restrictions on the frequency or quantity of content collection. The crawling method is used to collect data from blogs, forums, reviews, news, and YouTube, where streaming and pinging collection methods are not available. Crimson Hexagon partners with industry-leading third party content crawlers to provide data from the aforementioned sources. Collection software, provided by the third party, searches for new content that has been added to a website or blog, collects that content, and then sends it to Crimson Hexagon for analysis. Crimson Hexagon does not have an exhaustive list of all Internet data in their content library, but it is continually expanding the list of sites and sources. Currently, Crimson Hexagon provides blog content from July 2008 to present day at a rate of 1,100,000 posts per day; forum content from October 2008 to present day at a rate of 11,000,000 posts per day; news content from June 2008 to present day at a rate of 250,000 posts per day; and comments content from July 2010 to present day at a rate of 150,000 per day. Figure 1 describes the process of research, from the creation of the initial seed list of terms and the creation of high-value search terms, to the analysis of the online conversation, to the identification of social and digital influencers, and finally the development of insights and recommendations to improve content reach and resonance.
450 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Figure 1 Digital and Social Media Research Process SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
Defining the Conversation: Keyword Search First, to determine whether the bullying prevention content created and distributed by the Federal Partners in Bullying Prevention through StopBullying.gov is likely to reach target audiences who seek anti-bullying resources, Google search data were used to examine what terms Internet users were using to find information. Google search data are the largest available data set, and they were used to provide insight into terms and related topics in online searches related to bullying. A short but inclusive list of all related keywords and terms, known as a seed list, was created. The seed list was then expanded by calculating all related search terminology in Google’s database. The final term list was sorted and filtered by usage volume, eliminating “noise” that contains high volume but low relevance, and resulted in a list of several hundred terms and phrases. That comprehensive list was narrowed down to a list of the 84 highest-value keywords, as defined by search volume, competition, and the term’s relevance to StopBullying.gov’s objectives and priorities. Table 1 gives a clear picture of the most used and valued terminology for the topic area. The initial investigation included hundreds of words and phrases related to bullying. Close examination of the
results revealed a much smaller subset of words and phrases that were most commonly used in searches. These frequently used search terms were then compared with the language used on StopBullying.gov and its companion social media streams. Keywords and terms were then distilled to identify terms that were high value— those that had both strong search request volume and drove high web traffic (website visits). “Noisy” terms that had high request volume but were off-topic or did not drive traffic were filtered out. For example, “workplace bully” is not relevant to StopBullying.gov’s work and would be classified as noisy. The resulting key terms index list became the foundation for the following research steps. A monthly analysis of Google keyword searches is ongoing to ensure that HRSA remains up to date on the language of the conversation. Research and Analysis: Listening to the Conversation Following the discovery of what terms people used to seek information about bullying, the team explored what people were saying about the topic across platforms, who was leading the conversations, and what the tone of their commentary was. Two areas of research were emphasized:
Edgerton et al. / New Strategies to Assess Online Health Communication 451
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Table 1 Most Frequently Used Internet Search Terms for Bullying (November 2, 2012 to January 17, 2013) what is a bully bullying what is bullying about bullying the bullying bullies what are bullies anti bullying bullys statistics on bullying school bullying bullying statistics statistics about bullying bullying in school statistics for bullying statistics of bullying bullying at school what is bullying in school statistics bullying bullying school in school bullying bullying in the school
6,120,000 5,000,000 5,000,000 5,000,000 5,000,000 4,090,000 4,090,000 201,000 201,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000
SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
••
••
Conversation Analysis, including overall web traffic, demographic information (sex, age, and geographic location when available) and the following: || Sentiment trends—whether conversations were positive, negative, or neutral in nature || Conversation volume—the number of participants in a conversation and its fluctuation over time || Channels—where the conversation was taking place (e.g. Twitter, Facebook, blogs, news outlets such as The Charlotte Observer and blogs such as cafemom.com and bigeducationape. com) || Subtopics—specific subjects of conversation within each broad topic Influencer Identification and Analysis, focusing on certain individuals who command the attention of social media users and whose posts on
bullying and bullying prevention gain significant attention and reach.
To carry out the conversation analysis, a tool called Crimson Hexagon was used to map where current conversations about bullying take place online. The initial key terms index list was used to define the conversation and then the tool was trained to determine if conversations during a six-month period were categorized as very positive/solutions-oriented, leaning positive/ relevant, negative, or neutral. These categorization terms (positive/relevant, negative, and neutral) were developed for this study as a way to interpret data. “Positive” was defined as solutions-oriented, sharing research and best practices, or encouraging others to prevent bullying in ways that are safe and effective. “Leaning Positive” was defined as positive, but not especially helpful, for example, “#StopBullying.” “Negative” was defined as snide comments, profanity, angry or inciting language, or encouragement of others to address bullying in unsafe ways. “Neutral” referred to experience-based commentary, a balance of positive and negative ideas, or the use of bullying in a different sense, such as “The president is bullying Congress.” Large quantities of online content, including blogs, tweets, Facebook updates, and online news articles, were mapped in the analysis to ensure deep insight into the quality of the conversation and the accuracy of the information being shared. Next, to assess the identity and reach of key online influencers, an online tool called Traackr was used to measure the following three metrics: •• •• ••
Reach—measure of total audience size. Key indicators include: blog visitors, Twitter following, and YouTube subscribers. Resonance—measure of how much activity someone creates when he or she publishes. Key indicators include: Twitter retweets, linkbacks, and comments. Relevance—measure of how prolific someone is on a topic. Key indicators include: frequency, timing, diversity, and placement of keywords.
Our research process (see Figure 1) appears to be unique in scope, as the metrics used for measuring influence and maximizing the capability of social media as a public health tool have not yet been applied to online conversations about bullying. It is helpful for understanding the issue at both a macro- and microlevel, offering insight into what is driving the discussion and who is participating.
452 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Figure 2 Daily Bullying Opinion Analysis and Volume Trends per Day (November 2, 2012 to January 17, 2013) SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
Results >> Key findings from the keyword research, conversation analysis, and influencer analysis included the following:
1. Seven search term combinations represented the vast majority of bullying enquiries. These term combinations included “what is a bully,” “what is bullying,” and “about bullying,” indicating that adults and children look for information and resources about the bullying definition most frequently. Together, these terms accounted for more than 34 million monthly searches (Table 1). 2. A significant gap existed between language individuals use when searching for information about bullying and vocabulary that content providers such as StopBullying.gov use when discussing the same topic. Terms such as “bullying prevention and response,” which are common among public health practitioners, were not used in searches.
3. Most conversations about bullying were categorized as positive or very positive, indicating they were aligned with what HRSA and the Federal Partners in Bullying Prevention were saying about the issue (see Figure 2, which provides a visual of positive, negative/not helpful, and neutral conversations about bullying that occurred each day between November 2, 2012 and January 17, 2013.). We found that, of the 1,356,837 individual social or digital conversations analyzed, 33% of the conversations were very positive or solutions-oriented, 29% were leaning positive or relevant, 15% were negative/not helpful, and 22% were neutral. Many of the conversations that were positive included hashtag campaigns to drive specific antibullying awareness or prevention-oriented content and advice. The majority of the negative conversations included violent, profane, and/or disparaging language about individuals or antibullying efforts. 4. An analysis of channels (see Figure 3) indicated that Twitter hosted the largest share of bullying conversations. It housed approximately 41% of the conversation, followed by news websites (18%)
Edgerton et al. / New Strategies to Assess Online Health Communication 453
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
and blogs (19%). Public conversations on Facebook accounted for only 8% of the discussion. 5. Visuals were the most successful type of content. Results showed that it was critical to share the same information in a variety of visual formats across different channels to reach and resonate with broad audiences. Research showed that the most popular types of content were infographics, “Top 10” lists, “Myth vs. Fact” posts, videos, and images. 6. Several highly influential individuals and organizations were at the forefront of the bullying conversation. They included prominent bloggers, researchers, and organizations that consistently post or tweet about bullying, have a high number of followers, and whose content is shared widely.
Discussion >> Results of the analysis have been used to develop recommendations and comprehensive ideas for improving the reach and resonance of StopBullying.gov content and communications. They have assisted in driving traffic to the website and improving the content, so users will like, share, retweet, comment, and use online information to develop offline activities such as bullying prevention training or successful intervention in bullying situations. Bridging the Gap Between Practitioners and the Public When creating public health campaigns or resources, it is more valuable to discover and analyze what users demonstrate as relevant through searches than to start with an agency’s own internal terms. In this case, there were seven search term combinations that represented the vast majority of bullying enquiries, and much of the public health language used by coalition partners was not used by lay audiences in searching. Using the language that members of the target audience are using is key to reaching them, connecting with them, and ensuring messages resonate. For example, parents conducting online searches are unlikely to use social science or research terminology. They do not describe their children as “adolescents” or “youth” but as “kids” or “teenagers.” They seek information on ways to “stop bullying,” not on “bullying prevention and response.” Occasionally, they may misspell words, so including misspellings of search terms in the analysis—such as “bullys” instead of “bullies”—was also necessary to make the connection between what is being searched for and available resources.
Figure 3 Bullying Conversation Channel Breakdown (November 2, 2012 to January 17, 2013) SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
Since completing the analysis, HRSA and Widmeyer have used these insights and others to develop specific content recommendations for the language used on the website and social media platforms. Where appropriate, “kids” has been substituted for “children and youth” and “stop bullying” has been used instead of “bullying prevention and response.” Because “statistics about bullying” was a top search term, the team reviewed relevant statistical content on StopBullying. gov and identified pages where it was helpful to incorporate metadata keywords in order to attract those searching for this content. Also, because infographics were consistently popular across audiences and likely to be shared, the team has focused on connecting searchers seeking statistics with infographics containing key bullying statistics. Remembering the Social in Social Media When individuals and organizations capture a significant amount of attention in social media conversation on a particular topic, they are deemed influencers. These individuals are important bellwethers of the discussion, which can change quickly. It is also easier to cultivate relationships with a finite number of social media participants than trying to cast an indefinite net. To capitalize on the conversations already taking place and to take advantage of influencers’ authority, HRSA and Widmeyer have used influencer data and analysis to inform the following tactics:
454 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
•• •• ••
Connect with top online influencers by following them on Twitter and liking their organizations on Facebook Retweet, share, and comment regularly on influencers’ activity that aligns with federal messages Share StopBullying.gov content and Tumblr posts directly with influencers, utilizing terms and hashtags they use frequently
Liking and following organizations on Facebook and Twitter has already contributed to stronger engagement with key influencers, who frequently share and comment on StopBullying.gov activity. Prior to the analysis, the StopBullying.gov Twitter platform followed only 19 profiles or “handles” (all federal organizations or individuals) and liked zero Facebook pages. Today, StopBullying.gov follows and engages with 200 handles and likes 159 pages on Facebook. By regularly compiling a list of the current top influencers and examining their role in the conversation, the team has been able to go beyond standard metrics of success and assess the larger online conversation in bullying prevention against the chatter that fans and followers engaged in on StopBullying.gov’s social media platforms. HRSA and Widmeyer have been able to set measurable goals for growth and monitor against those goals on a monthly and quarterly basis. Another benefit is the opportunity to use the data to create new and better social media content. Understanding what type of content generates the most traction is valuable for creating new content and developing more engaging illustrations of critical information—for influencers and mass audiences alike. To further maximize the reach of StopBullying.gov and exposure to its content, other recommendations have included the following: •• •• •• •• ••
Incorporate high-value keywords into StopBullying. gov content and metadata to improve search performance, expand reach, and engage audiences Monitor and mirror influencers’ vocabulary Increase cross-promotion of content across all social channels Live-tweet from conferences and events Increase participation in relevant Twitter chats to share resources and increase engagement
Tracking and Reporting on a Broad Set of Indicators In addition to tracking the basic statistics, such as followers on social media, the team has looked at a wide variety of metrics to understand the tone and direction of the conversation and align StopBullying.gov content accordingly. New metrics include the following:
•• •• •• •• •• •• •• •• •• •• ••
Stories created (includes total likes, comments, and shares) Click-through rates Engagement rates Audience demographics (includes age, gender, and location) Top posts Top terms shared by influencers Top links shared by influencers Reach score Resonance score Relevance score Klout score
Conclusion >> Social media, when done right, holds great promise for public health practitioners. As the adoption of Facebook, Twitter, and other online engagement tools becomes more common among leaders of the field, so too will the insights and sharing of best practices. Guided by an initial interest in contextualizing social media successes and demonstrating return on investment for its digital communications efforts, the team developed a methodology that is applicable to a wide range of public health issues, from smoking cessation to healthy eating. While the long-term results of this effort continue to unfold, the research and analysis have played a crucial role in the team’s ability to develop recommendations and comprehensive ideas for maximizing the impact of StopBullying.gov content and communications. Since completing the analysis, HRSA and Widmeyer have successfully expanded the reach and resonance of the website’s social media platforms—doubling its Twitter following in 2013, organically growing its Facebook page by more than 50,000 likes, and generating 55 million page impressions across the two platforms. Leveraging new connections with more than 200 influencers and organizations, the team has been able to amplify key messages about the importance of bullying prevention; promote new resources; and share outside information generated by experts and leading stakeholders such as bloggers, journalists, and researchers. HRSA is now better positioned to provide target audiences with research-based materials about bullying, educate stakeholders about the public health effects of bullying, and ultimately decrease the likelihood that children will be bullied in the future. References Bogart, L. M., Elliott, M. N., Klein, D. J., Tortolero, S. R., Mrug, S., Peskin, M. F., . . . Schuster, M. A. (2014). Peer victimization in 5th grade and health in 10th grade. Pediatrics, 133, 440-447.
Edgerton et al. / New Strategies to Assess Online Health Communication 455
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32, 554-571. Espelage, D. L., & Swearer, S. M. (2010). A social-ecological model for bullying prevention and intervention: Understanding the impact of adults in the social ecology of youngsters. In S. R. Jimerson, S. M. Swearer, & D. L. Espelage (Eds.), Handbook of bullying in schools: An international perspective (pp. 61-72). New York, NY: Routledge. Gini, G., & Pozzoli, T. (2013). Bullied children and psychosomatic problems: A meta-analysis. Pediatrics, 132, 720-729. Gladden, R. M., Vivolo-Kantor, A. M., Hamburger, M. E., & Lumpkin, C. D. (2014). Bullying surveillance among youths: Uniform definitions for public health and recommended data elements, Version 1.0. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, and U.S. Department of Education. Harris, J. K., Mueller, N. L., & Snider, D. (2013). Social media adoption in local health departments nationwide. American Journal of Public Health, 103, 1700-1707. Kahai, S. S., & Cooper, R. B. (2003). Exploring the core concepts of media richness theory: The impact of cue multiplicity and feedback immediacy on decision quality. Journal of Management Information Systems, 20, 263-299. Ledford, C. J. W. (2012). Changing channels: A theory-based guide to selecting traditional, new, and social media in strategic social marketing. Social Marketing Quarterly, 18, 175-186. Lilley, R. (2007). What can social marketing offer public health? Primary Health Care, 17(9), 22-23. Long, T., Taubenheim, A. M., Wayman, J., Temple, S., & Yu, E. (2010). Using social media and internet marketing to reach
women with The Heart Truth. Cases in Public Health Communication & Marketing, 4, 55-68. Retrieved from http:// www.nhlbi.nih.gov/educational/hearttruth/downloads/html/smstudy.htm Neiger, B. L., Thackeray, R., Burton, S. H., Giraud-Carrier, C. G., & Fagen, M. C. (2013). Evaluating social media’s capacity to develop engaged audiences in health promotion settings: Use of twitter metrics as a case study. Health Promotion Practice, 14, 157-162. Robers, S., Zhang, A., Morgan, R. E., & Musu-Gillette, L. (2015). Indicators of school crime and safety: 2014 (NCES 2015-072/NCJ 248036). Washington, DC: National Center for Education Statistics, U.S. Department of Education, and Bureau of Justice Statistics, Office of Justice Programs, U.S. Department of Justice. Thackeray, R., Neiger, B. L., Burton, S. H., & Thackeray, C. R. (2013). Analysis of the purpose of state health departments’ tweets: Information sharing, engagement, and action. Journal of Medical Internet Research, 15(11), e255. doi:10.2196/jmir.3002. Retrieved from http://www.jmir.org/2013/11/e255/ Thackeray, R., Neiger, B. L., Smith, A. K., & Van Wagenen, S. B. (2012). Adoption and use of social media among public health departments. BMC Public Health, 12(242), 1-6. Ttofi, M. M., Farrington, D. P., Lösel, F., & Loeber, R. (2011a). Do the victims of school bullies tend to become depressed later in life? A systematic review and meta-analysis of longitudinal studies. Journal of Aggression, Conflict and Peace Research, 3, 63-73. Ttofi, M. M., Farrington, D. P., Lösel, F., & Loeber, R. (2011b). The predictive efficiency of school bullying versus later offending: A systematic/meta-analytic review of longitudinal studies. Criminal Behaviour and Mental Health, 21, 80-89.
456 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
HPPXXX10.1177/1524839915620392Health Promotion PracticeEdgerton et al. / New Strategies to Assess Online Health Communication
Social Media & Community-Engaged Practices
Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention Elizabeth Edgerton, MD, MPH1 Erin Reiney, MPH, CHES1 Siobhan Mueller2 Barry Reicherter2 Katherine Curtis2 Stephanie Waties2 Susan P. Limber, PhD3
Every day in classrooms, playgrounds and school hallways, through text messages and mobile technology apps, children are bullied by other children. Conversations about this bullying—what it is, who is involved, and how to stop it—are taking place online. To fill a need for relevant, research-based materials on bullying, the U.S. Department of Health and Human Services’ Health Resources and Services Administration worked with Widmeyer Communications to investigate the scope of media conversations about bullying and discover new strategies for promoting appropriate public health messages about bullying to intended audiences. Key components of the methodology included: analyzing common search terms and aligning social media content with terms used in searches rather than technical language; identifying influencers in social media spheres, cultivating relationships with them, and sharing their positive, relevant content; examining which digital formats are most popular for sharing and creating content across platforms; tracking and reporting on a wide variety of metrics (such as click-through and engagement rates and reach, resonance, relevance, and Klout scores) to understand conversations around bullying; and looking at online conversations and engaging participants using applicable resources and calls to action. A key finding included a significant gap
Health Promotion Practice May 2016 Vol. 17, No. (3) 448–456 DOI: 10.1177/1524839915620392 © 2016 Society for Public Health Education
between search terms and online content and has led to recommendations and comprehensive ideas for improving the reach and resonance of StopBullying.gov content and communications. Keywords: social marketing; social media; health communication; metrics and evaluation; bullying prevention; online influencer identification; keyword research; conversation mapping
B
ullying is an all too common experience in the United States, with more than one in five youth between the ages of 12 and 18 years targeted at school (Robers, Zhang, Morgan, & Musu-Gillette, 2015). It is defined as unwanted aggressive behavior among school-aged children that “involves an observed or perceived power imbalance and is repeated multiple times or is highly likely to be repeated” (Gladden, VivoloKantor, Hamburger, & Lumpkin, 2014, p. 7). Longitudinal research has found that chronic bullying impairs victims’ health, but identification and early intervention 1
Health & Human Services, Health Resources and Services Administration, Rockville, MD, USA 2 Widmeyer Communications, Washington, DC, USA 3 Clemson University, Clemson, SC, USA Authors’ Note: Address correspondence to Siobhan Mueller, Widmeyer Communications, 1129 20th Street NW, Suite 200, Washington, DC, USA; e-mail: [email protected].
448
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
by clinicians can stem negative effects (Bogart et al., 2014; Gini & Pozzoli, 2013). Because of the high prevalence of bullying and the serious short- and long-term effects identified in children who are bullied (Ttofi, Farrington, Lösel, & Loeber, 2011a), children who bully others (Ttofi, Farrington, Lösel, & Loeber, 2011b), and even those who witness the behavior (Espelage & Swearer, 2010), bullying is recognized as a public health issue. In 2001, the Heath Resources and Services Administration (HRSA) joined efforts to address bullying in schools and communities by launching a social marketing and communications campaign to promote prevention and intervention strategies. HRSA selected Widmeyer Communications to facilitate the planning, design, production, and dissemination of the initial campaign materials and continues to develop content and support communication strategies. In 2011, HRSA joined forces with other government agencies to coordinate efforts and establish the Federal Partners in Bullying Prevention, which includes the departments of Education, Defense, and Justice, among others. The Federal Partners launched StopBullying.gov, a website with research, best practices, and training materials for bullying prevention and response. Stakeholders include, but are not limited to: parents, educators, health and safety professionals, law enforcement officers, and child-care providers. Social media messaging was also introduced through a Facebook page and Twitter profile to complement the main portal of web-based information at StopBullying.gov by bringing attention to the site and promoting its content. The goals of these digital assets were to expand the number of community stakeholders being reached, and to further connect the information and best practices available on the website with the target audiences. Traditional metrics such as number of site visits, Twitter followers, and Facebook likes were used to track progress over time. Data indicated that StopBullying.gov platforms experienced consistent but moderate growth between 2011 and 2013; however these indicators were limited because they only showed numbers of people who used online resources. They did not accurately indicate whether people who searched for help addressing bullying found the information they needed from the site or its social media pages.
Literature Review >> In the past, using social media to promote public health campaigns or other socially beneficial concepts has been primarily restricted to pushing out untailored, one-way messages (Neiger, Thackeray, Burton, GiraudCarrier, & Fagen, 2013). Few organizations have mastered
the art of meaningful exchanges with constituents or consumers, and as a result, they fail to take advantage of the interactive nature of social media. For many social media managers with government agencies, the slow pace of social media adaptation does not stem from a lack of understanding or desire, but from restrictive policies imposed on content and postings. Media richness theory proposes that media differ in their abilities to facilitate changes in understanding among communicators (Daft & Lengel, 1986; Kahai & Cooper, 2003). As Ledford (2012) notes, social media channels rank high on both richness (quick response, ability to communicate multiple cues, ability for natural language, and personal focus) and content control. Because of content control challenges, the nature of social media by public agencies tends to focus on developing and sharing their own content, responding to comments or questions, sharing content from nongovernmental sources, and retweeting content or links from private citizens. Comments and feedback are moderated and shared content is controlled, lacking true engagement (Thackeray, Neiger, Burton, & Thackeray, 2013). The potential of social media as a meaningful influencer of public health promotion is evident in the numbers: Researchers estimate that by 2017, the number of individuals and corporations who have social networking accounts will reach more than 4.8 billion. In particular, Twitter is becoming ever more popular. A total of 18% of Internet users have a Twitter account . . . [It] has been used to disseminate information about diabetes, breast cancer, to communicate during a disaster, and to understand health-related trends and issues such as influenza, tobacco, prescription drug misuse, and suicide. (Thackeray et al., 2013)
Still, social platforms are not yet being used to their full advantage by health departments (Thackeray et al., 2013) and other public agencies. The most successful social marketing campaigns leverage social media to its full capacity. A good example is The Heart Truth, the first national awareness campaign designed to educate women about their risks of heart disease, the leading cause of death among U.S. women. Sponsored by the Health and Human Services’ National Heart, Lung, and Blood Institute, the campaign harnessed social media marketing and digital strategies in the campaign’s design and in the highly successful Red Dress brand that continues to be used 5 years later by the National Wear Red Day and the Red Dress Collection (Long et al., 2009). According to Long et al. (2009), The Heart
Edgerton et al. / New Strategies to Assess Online Health Communication 449
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Truth campaign leveraged a range of social platforms and applications, including blogs, Facebook, YouTube, Wikipedia, and Flickr, as well as online public service announcements and pay-per-click ads. Targeted outreach was also conducted to influential bloggers who write about women, health, motherhood, fashion, and entertainment. The Heart Truth campaign is credited with having increased women’s awareness about heart disease and its prevention and contributed to a reduction in rates between 2002 and 2008 (Long et al., 2009). Lilley (2007) describes the potential of applying commercial marketing techniques to the social sector in this way: Just as a large chain store tracks consumers’ purchases and predicts future needs as a result, public health departments could target individuals interested in smoking cessation or weight loss with appropriate messages and resources. A study by Harris, Mueller, and Snider (2013), which examined the use of social media by local health departments in promoting health messages, concluded the practice is not widespread, but it is emerging especially in large metropolitan areas. A cross-sectional study of state public health departments by Thackeray, Neiger, Smith, and Van Wagenen (2012) confirmed that adoption of social media by these agencies is still in the early stage and primarily used as a digital version of traditional oneway dissemination. Most health departments are not optimizing the potential of social media platforms for interactivity and engagement that could result in improved health outcomes (Thackeray et al., 2012). Developing customized digital strategies for using, evaluating, and revising social media activities that are engaging and interactive is still very unique. Moreover, to our knowledge, no studies of social media strategies for bullying prevention have been published to date. Recognizing these gaps in the field, the purpose of this study was several fold: (a) investigate the scope of social media conversations around a critical public health issue—bullying, including reach resonance, and relevance; (b) discover new strategies to leverage social media more effectively for the promotion of research-based public health messages about bullying and best practices in bullying prevention; and (c) determine the extent to which this content is likely to reach new key target audiences.
Method >> In 2013, HRSA collaborated with Widmeyer Communications to conduct this research which included the collection of data from 2011-2012. The research approach was divided into two broad categories:
••
••
Defining the conversation by identifying relevant audiences and topics, creating a seed term list with popular search terms and keywords, and then creating a key search term list based on examining Google search data. For the purposes of this research effort, conversation is defined as any public social or digital media communication about bullying that falls within the keyword query parameters. The conversation was analyzed on Twitter, Facebook, blogs, discussion forums, news sites, and YouTube. The conversation is composed of individual tweets, blogs, articles, and comments shared on the aforementioned platforms. Comments, which can be left on news articles and blogs, are defined as their own conversation category in the analysis. Research and analysis consisting of conversation targeting, monitoring, and data collection; analysis; and message creation. The tool used to conduct the conversation analysis, Crimson Hexagon, collects data through three methods: streaming, pinging, and crawling. Streaming data is the most direct method and provides real-time, unrestricted access to the Twitter content stream. Pinging content collection, or API (application programming interface) call, is used on Facebook, Instagram, and Google+. Pinging also provides direct access to platform data, but the source may place restrictions on the frequency or quantity of content collection. The crawling method is used to collect data from blogs, forums, reviews, news, and YouTube, where streaming and pinging collection methods are not available. Crimson Hexagon partners with industry-leading third party content crawlers to provide data from the aforementioned sources. Collection software, provided by the third party, searches for new content that has been added to a website or blog, collects that content, and then sends it to Crimson Hexagon for analysis. Crimson Hexagon does not have an exhaustive list of all Internet data in their content library, but it is continually expanding the list of sites and sources. Currently, Crimson Hexagon provides blog content from July 2008 to present day at a rate of 1,100,000 posts per day; forum content from October 2008 to present day at a rate of 11,000,000 posts per day; news content from June 2008 to present day at a rate of 250,000 posts per day; and comments content from July 2010 to present day at a rate of 150,000 per day. Figure 1 describes the process of research, from the creation of the initial seed list of terms and the creation of high-value search terms, to the analysis of the online conversation, to the identification of social and digital influencers, and finally the development of insights and recommendations to improve content reach and resonance.
450 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Figure 1 Digital and Social Media Research Process SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
Defining the Conversation: Keyword Search First, to determine whether the bullying prevention content created and distributed by the Federal Partners in Bullying Prevention through StopBullying.gov is likely to reach target audiences who seek anti-bullying resources, Google search data were used to examine what terms Internet users were using to find information. Google search data are the largest available data set, and they were used to provide insight into terms and related topics in online searches related to bullying. A short but inclusive list of all related keywords and terms, known as a seed list, was created. The seed list was then expanded by calculating all related search terminology in Google’s database. The final term list was sorted and filtered by usage volume, eliminating “noise” that contains high volume but low relevance, and resulted in a list of several hundred terms and phrases. That comprehensive list was narrowed down to a list of the 84 highest-value keywords, as defined by search volume, competition, and the term’s relevance to StopBullying.gov’s objectives and priorities. Table 1 gives a clear picture of the most used and valued terminology for the topic area. The initial investigation included hundreds of words and phrases related to bullying. Close examination of the
results revealed a much smaller subset of words and phrases that were most commonly used in searches. These frequently used search terms were then compared with the language used on StopBullying.gov and its companion social media streams. Keywords and terms were then distilled to identify terms that were high value— those that had both strong search request volume and drove high web traffic (website visits). “Noisy” terms that had high request volume but were off-topic or did not drive traffic were filtered out. For example, “workplace bully” is not relevant to StopBullying.gov’s work and would be classified as noisy. The resulting key terms index list became the foundation for the following research steps. A monthly analysis of Google keyword searches is ongoing to ensure that HRSA remains up to date on the language of the conversation. Research and Analysis: Listening to the Conversation Following the discovery of what terms people used to seek information about bullying, the team explored what people were saying about the topic across platforms, who was leading the conversations, and what the tone of their commentary was. Two areas of research were emphasized:
Edgerton et al. / New Strategies to Assess Online Health Communication 451
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Table 1 Most Frequently Used Internet Search Terms for Bullying (November 2, 2012 to January 17, 2013) what is a bully bullying what is bullying about bullying the bullying bullies what are bullies anti bullying bullys statistics on bullying school bullying bullying statistics statistics about bullying bullying in school statistics for bullying statistics of bullying bullying at school what is bullying in school statistics bullying bullying school in school bullying bullying in the school
6,120,000 5,000,000 5,000,000 5,000,000 5,000,000 4,090,000 4,090,000 201,000 201,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000 165,000
SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
••
••
Conversation Analysis, including overall web traffic, demographic information (sex, age, and geographic location when available) and the following: || Sentiment trends—whether conversations were positive, negative, or neutral in nature || Conversation volume—the number of participants in a conversation and its fluctuation over time || Channels—where the conversation was taking place (e.g. Twitter, Facebook, blogs, news outlets such as The Charlotte Observer and blogs such as cafemom.com and bigeducationape. com) || Subtopics—specific subjects of conversation within each broad topic Influencer Identification and Analysis, focusing on certain individuals who command the attention of social media users and whose posts on
bullying and bullying prevention gain significant attention and reach.
To carry out the conversation analysis, a tool called Crimson Hexagon was used to map where current conversations about bullying take place online. The initial key terms index list was used to define the conversation and then the tool was trained to determine if conversations during a six-month period were categorized as very positive/solutions-oriented, leaning positive/ relevant, negative, or neutral. These categorization terms (positive/relevant, negative, and neutral) were developed for this study as a way to interpret data. “Positive” was defined as solutions-oriented, sharing research and best practices, or encouraging others to prevent bullying in ways that are safe and effective. “Leaning Positive” was defined as positive, but not especially helpful, for example, “#StopBullying.” “Negative” was defined as snide comments, profanity, angry or inciting language, or encouragement of others to address bullying in unsafe ways. “Neutral” referred to experience-based commentary, a balance of positive and negative ideas, or the use of bullying in a different sense, such as “The president is bullying Congress.” Large quantities of online content, including blogs, tweets, Facebook updates, and online news articles, were mapped in the analysis to ensure deep insight into the quality of the conversation and the accuracy of the information being shared. Next, to assess the identity and reach of key online influencers, an online tool called Traackr was used to measure the following three metrics: •• •• ••
Reach—measure of total audience size. Key indicators include: blog visitors, Twitter following, and YouTube subscribers. Resonance—measure of how much activity someone creates when he or she publishes. Key indicators include: Twitter retweets, linkbacks, and comments. Relevance—measure of how prolific someone is on a topic. Key indicators include: frequency, timing, diversity, and placement of keywords.
Our research process (see Figure 1) appears to be unique in scope, as the metrics used for measuring influence and maximizing the capability of social media as a public health tool have not yet been applied to online conversations about bullying. It is helpful for understanding the issue at both a macro- and microlevel, offering insight into what is driving the discussion and who is participating.
452 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Figure 2 Daily Bullying Opinion Analysis and Volume Trends per Day (November 2, 2012 to January 17, 2013) SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
Results >> Key findings from the keyword research, conversation analysis, and influencer analysis included the following:
1. Seven search term combinations represented the vast majority of bullying enquiries. These term combinations included “what is a bully,” “what is bullying,” and “about bullying,” indicating that adults and children look for information and resources about the bullying definition most frequently. Together, these terms accounted for more than 34 million monthly searches (Table 1). 2. A significant gap existed between language individuals use when searching for information about bullying and vocabulary that content providers such as StopBullying.gov use when discussing the same topic. Terms such as “bullying prevention and response,” which are common among public health practitioners, were not used in searches.
3. Most conversations about bullying were categorized as positive or very positive, indicating they were aligned with what HRSA and the Federal Partners in Bullying Prevention were saying about the issue (see Figure 2, which provides a visual of positive, negative/not helpful, and neutral conversations about bullying that occurred each day between November 2, 2012 and January 17, 2013.). We found that, of the 1,356,837 individual social or digital conversations analyzed, 33% of the conversations were very positive or solutions-oriented, 29% were leaning positive or relevant, 15% were negative/not helpful, and 22% were neutral. Many of the conversations that were positive included hashtag campaigns to drive specific antibullying awareness or prevention-oriented content and advice. The majority of the negative conversations included violent, profane, and/or disparaging language about individuals or antibullying efforts. 4. An analysis of channels (see Figure 3) indicated that Twitter hosted the largest share of bullying conversations. It housed approximately 41% of the conversation, followed by news websites (18%)
Edgerton et al. / New Strategies to Assess Online Health Communication 453
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
and blogs (19%). Public conversations on Facebook accounted for only 8% of the discussion. 5. Visuals were the most successful type of content. Results showed that it was critical to share the same information in a variety of visual formats across different channels to reach and resonate with broad audiences. Research showed that the most popular types of content were infographics, “Top 10” lists, “Myth vs. Fact” posts, videos, and images. 6. Several highly influential individuals and organizations were at the forefront of the bullying conversation. They included prominent bloggers, researchers, and organizations that consistently post or tweet about bullying, have a high number of followers, and whose content is shared widely.
Discussion >> Results of the analysis have been used to develop recommendations and comprehensive ideas for improving the reach and resonance of StopBullying.gov content and communications. They have assisted in driving traffic to the website and improving the content, so users will like, share, retweet, comment, and use online information to develop offline activities such as bullying prevention training or successful intervention in bullying situations. Bridging the Gap Between Practitioners and the Public When creating public health campaigns or resources, it is more valuable to discover and analyze what users demonstrate as relevant through searches than to start with an agency’s own internal terms. In this case, there were seven search term combinations that represented the vast majority of bullying enquiries, and much of the public health language used by coalition partners was not used by lay audiences in searching. Using the language that members of the target audience are using is key to reaching them, connecting with them, and ensuring messages resonate. For example, parents conducting online searches are unlikely to use social science or research terminology. They do not describe their children as “adolescents” or “youth” but as “kids” or “teenagers.” They seek information on ways to “stop bullying,” not on “bullying prevention and response.” Occasionally, they may misspell words, so including misspellings of search terms in the analysis—such as “bullys” instead of “bullies”—was also necessary to make the connection between what is being searched for and available resources.
Figure 3 Bullying Conversation Channel Breakdown (November 2, 2012 to January 17, 2013) SOURCE: U.S. Department of Health and Human Services, Health Resources and Services Administration, Identifying New Strategies to Assess and Promote Online Health Communication and Social Media Outreach: An Application in Bullying Prevention (2015).
Since completing the analysis, HRSA and Widmeyer have used these insights and others to develop specific content recommendations for the language used on the website and social media platforms. Where appropriate, “kids” has been substituted for “children and youth” and “stop bullying” has been used instead of “bullying prevention and response.” Because “statistics about bullying” was a top search term, the team reviewed relevant statistical content on StopBullying. gov and identified pages where it was helpful to incorporate metadata keywords in order to attract those searching for this content. Also, because infographics were consistently popular across audiences and likely to be shared, the team has focused on connecting searchers seeking statistics with infographics containing key bullying statistics. Remembering the Social in Social Media When individuals and organizations capture a significant amount of attention in social media conversation on a particular topic, they are deemed influencers. These individuals are important bellwethers of the discussion, which can change quickly. It is also easier to cultivate relationships with a finite number of social media participants than trying to cast an indefinite net. To capitalize on the conversations already taking place and to take advantage of influencers’ authority, HRSA and Widmeyer have used influencer data and analysis to inform the following tactics:
454 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
•• •• ••
Connect with top online influencers by following them on Twitter and liking their organizations on Facebook Retweet, share, and comment regularly on influencers’ activity that aligns with federal messages Share StopBullying.gov content and Tumblr posts directly with influencers, utilizing terms and hashtags they use frequently
Liking and following organizations on Facebook and Twitter has already contributed to stronger engagement with key influencers, who frequently share and comment on StopBullying.gov activity. Prior to the analysis, the StopBullying.gov Twitter platform followed only 19 profiles or “handles” (all federal organizations or individuals) and liked zero Facebook pages. Today, StopBullying.gov follows and engages with 200 handles and likes 159 pages on Facebook. By regularly compiling a list of the current top influencers and examining their role in the conversation, the team has been able to go beyond standard metrics of success and assess the larger online conversation in bullying prevention against the chatter that fans and followers engaged in on StopBullying.gov’s social media platforms. HRSA and Widmeyer have been able to set measurable goals for growth and monitor against those goals on a monthly and quarterly basis. Another benefit is the opportunity to use the data to create new and better social media content. Understanding what type of content generates the most traction is valuable for creating new content and developing more engaging illustrations of critical information—for influencers and mass audiences alike. To further maximize the reach of StopBullying.gov and exposure to its content, other recommendations have included the following: •• •• •• •• ••
Incorporate high-value keywords into StopBullying. gov content and metadata to improve search performance, expand reach, and engage audiences Monitor and mirror influencers’ vocabulary Increase cross-promotion of content across all social channels Live-tweet from conferences and events Increase participation in relevant Twitter chats to share resources and increase engagement
Tracking and Reporting on a Broad Set of Indicators In addition to tracking the basic statistics, such as followers on social media, the team has looked at a wide variety of metrics to understand the tone and direction of the conversation and align StopBullying.gov content accordingly. New metrics include the following:
•• •• •• •• •• •• •• •• •• •• ••
Stories created (includes total likes, comments, and shares) Click-through rates Engagement rates Audience demographics (includes age, gender, and location) Top posts Top terms shared by influencers Top links shared by influencers Reach score Resonance score Relevance score Klout score
Conclusion >> Social media, when done right, holds great promise for public health practitioners. As the adoption of Facebook, Twitter, and other online engagement tools becomes more common among leaders of the field, so too will the insights and sharing of best practices. Guided by an initial interest in contextualizing social media successes and demonstrating return on investment for its digital communications efforts, the team developed a methodology that is applicable to a wide range of public health issues, from smoking cessation to healthy eating. While the long-term results of this effort continue to unfold, the research and analysis have played a crucial role in the team’s ability to develop recommendations and comprehensive ideas for maximizing the impact of StopBullying.gov content and communications. Since completing the analysis, HRSA and Widmeyer have successfully expanded the reach and resonance of the website’s social media platforms—doubling its Twitter following in 2013, organically growing its Facebook page by more than 50,000 likes, and generating 55 million page impressions across the two platforms. Leveraging new connections with more than 200 influencers and organizations, the team has been able to amplify key messages about the importance of bullying prevention; promote new resources; and share outside information generated by experts and leading stakeholders such as bloggers, journalists, and researchers. HRSA is now better positioned to provide target audiences with research-based materials about bullying, educate stakeholders about the public health effects of bullying, and ultimately decrease the likelihood that children will be bullied in the future. References Bogart, L. M., Elliott, M. N., Klein, D. J., Tortolero, S. R., Mrug, S., Peskin, M. F., . . . Schuster, M. A. (2014). Peer victimization in 5th grade and health in 10th grade. Pediatrics, 133, 440-447.
Edgerton et al. / New Strategies to Assess Online Health Communication 455
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media richness and structural design. Management Science, 32, 554-571. Espelage, D. L., & Swearer, S. M. (2010). A social-ecological model for bullying prevention and intervention: Understanding the impact of adults in the social ecology of youngsters. In S. R. Jimerson, S. M. Swearer, & D. L. Espelage (Eds.), Handbook of bullying in schools: An international perspective (pp. 61-72). New York, NY: Routledge. Gini, G., & Pozzoli, T. (2013). Bullied children and psychosomatic problems: A meta-analysis. Pediatrics, 132, 720-729. Gladden, R. M., Vivolo-Kantor, A. M., Hamburger, M. E., & Lumpkin, C. D. (2014). Bullying surveillance among youths: Uniform definitions for public health and recommended data elements, Version 1.0. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, and U.S. Department of Education. Harris, J. K., Mueller, N. L., & Snider, D. (2013). Social media adoption in local health departments nationwide. American Journal of Public Health, 103, 1700-1707. Kahai, S. S., & Cooper, R. B. (2003). Exploring the core concepts of media richness theory: The impact of cue multiplicity and feedback immediacy on decision quality. Journal of Management Information Systems, 20, 263-299. Ledford, C. J. W. (2012). Changing channels: A theory-based guide to selecting traditional, new, and social media in strategic social marketing. Social Marketing Quarterly, 18, 175-186. Lilley, R. (2007). What can social marketing offer public health? Primary Health Care, 17(9), 22-23. Long, T., Taubenheim, A. M., Wayman, J., Temple, S., & Yu, E. (2010). Using social media and internet marketing to reach
women with The Heart Truth. Cases in Public Health Communication & Marketing, 4, 55-68. Retrieved from http:// www.nhlbi.nih.gov/educational/hearttruth/downloads/html/smstudy.htm Neiger, B. L., Thackeray, R., Burton, S. H., Giraud-Carrier, C. G., & Fagen, M. C. (2013). Evaluating social media’s capacity to develop engaged audiences in health promotion settings: Use of twitter metrics as a case study. Health Promotion Practice, 14, 157-162. Robers, S., Zhang, A., Morgan, R. E., & Musu-Gillette, L. (2015). Indicators of school crime and safety: 2014 (NCES 2015-072/NCJ 248036). Washington, DC: National Center for Education Statistics, U.S. Department of Education, and Bureau of Justice Statistics, Office of Justice Programs, U.S. Department of Justice. Thackeray, R., Neiger, B. L., Burton, S. H., & Thackeray, C. R. (2013). Analysis of the purpose of state health departments’ tweets: Information sharing, engagement, and action. Journal of Medical Internet Research, 15(11), e255. doi:10.2196/jmir.3002. Retrieved from http://www.jmir.org/2013/11/e255/ Thackeray, R., Neiger, B. L., Smith, A. K., & Van Wagenen, S. B. (2012). Adoption and use of social media among public health departments. BMC Public Health, 12(242), 1-6. Ttofi, M. M., Farrington, D. P., Lösel, F., & Loeber, R. (2011a). Do the victims of school bullies tend to become depressed later in life? A systematic review and meta-analysis of longitudinal studies. Journal of Aggression, Conflict and Peace Research, 3, 63-73. Ttofi, M. M., Farrington, D. P., Lösel, F., & Loeber, R. (2011b). The predictive efficiency of school bullying versus later offending: A systematic/meta-analytic review of longitudinal studies. Criminal Behaviour and Mental Health, 21, 80-89.
456 HEALTH PROMOTION PRACTICE / May 2016
Downloaded from hpp.sagepub.com at University of Wollongong on May 17, 2016
