Using Digital and Social Media Metrics to Develop Mental Health Approaches for Youth PDF

Preview

Summary

Send Orders for Reprints to [email protected] Adolescent Psychiatry, 2014, 4, 000-000 1 Using Digital and Social Media Metrics to Develop Mental Health Approaches for Youth Christina Carew*, Stan Kutcher, Yifeng Wei and Alan McLuckie Dalhousie University & IWK Health Centre, Halifax, N...

Description

Accelerat ing t he world's research.

Using Digital and Social Media Metrics to Develop Mental Health Approaches for Youth alan mcluckie Adolescent Psychiatry

Cite this paper

Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers

Download a PDF Pack of t he best relat ed papers 

Teacher’s Percept ion About t he Use ofMet ilfenidat o for Brazilian Children at School Lisandre Frazão Brunelli Technophobia.docx Traci Belanger CPA Vict oria 2016 Symposium Faisal Hassan

Send Orders for Reprints to [email protected] Adolescent Psychiatry, 2014, 4, 000-000

1

Using Digital and Social Media Metrics to Develop Mental Health Approaches for Youth Christina Carew*, Stan Kutcher, Yifeng Wei and Alan McLuckie Dalhousie University & IWK Health Centre, Halifax, Nova Scotia, Canada Abstract: Objective: The objective of this project was to investigate the online behaviors of adolescent and adult populations with respect to mental health information seeking, and identify differences in approaches within age groups and geographical location. Method: This content analysis approach identified and mapped patterns in online conversations. The search data was able to quantify who was looking for teen mental health information, where they were looking, and what they were looking for. Additional analysis included the preferred format of information presentation and how mental health searches varied over time. Results: The results of the analysis revealed that between 2006 (baseline) to 2010, a 200% increase in online activity regarding mental health was identified. Adults were most likely to ignite (initiate) conversations online about depression, followed by: anxiety, doctors, suicide, treatments, and OCD. For teens, depression was also the most ignited topic area, followed by: anxiety, alcohol, suicide, sexting and marijuana. While adults were often seeking information about the disorders and treatment options, teens tended to discuss concerns through the use of personal stories. Conclusion: This research provides insight into how digital and social media can be used to engage both youth and adult discussions about mental health. We report substantive audience driven differences that can inform the development of targeted mental health knowledge translation methods and activities. A broader understanding of the key mental health topics of interest was garnered, in addition to how online use varied between audiences. These results have several implications for mental health knowledge translation including tactics to connect with various stakeholders.

Keywords: Digital media, knowledge translation, mental health, online behavior, online health promotion, social media, teens.. INTRODUCTION Digital and social media provides an accessible medium for individuals to seek and share information and has become an attractive tool for youth communications (Nie & Hillygus, 2002). In the healthcare profession, online tools are often used to disseminate information but are infrequently used as a means to better understand the target audience (Thackeray, Neiger, Hanson, & McKenzie, 2008). The role of electronic media in mental health knowledge translation has been relatively unexplored, however, “social media outlets may represent an excellent opportunity to reach traditionally underserved members of the population" (Wen-ying, Hunt, Beckjord, Moser, & Hesse, 2009). The Sun Life Financial Chair in Adolescent Mental Health Team is a knowledge translation group working in the area of adolescent mental health. In 2009, the team identified a need to better understand the audiences using their website and online resources. In particular, a focus was placed on understanding the needs/queries of these audiences related to *Address correspondence to this author at: Sun Life Financial Chair in Adolescent Mental Health, Dalhousie University & IWK Health Centre, 5850 University Avenue, PO Box 9700, Halifax, Nova Scotia, Canada B3K 6R8; Tel: 902-470-6598; Fax: 902-492-0383; E-mail: [email protected] 2210-6766/14 $58.00+.00

adolescent mental health. The group wanted to know if existing resources and information met the needs of individuals looking for mental health information, and had an interest in identifying areas that could be further developed. Previous attempts to determine the adolescent mental health information needs of parents, teens, educators and health professionals have relied on the use of qualitative approaches including focus groups and consultations. Focus groups allow participants to provide anecdotal information about their online behaviors and habits, however, they have the potential for bias. While these qualitative approaches were helpful, the team wanted to track the actual behaviors of the target audience in the same medium they wanted to attract them - online. The research method employed tracks online actions and findings about genuine behavior. An added benefit to this approach is that it provides an opportunity to dramatically increase the amount of data gathered without incurring increased time or cost commitments. The objective of this project was to investigate the online behaviors of adolescent and adult populations with respect to mental health information by conducting content analysis of data that tracked digital and social media habits. The information sought after included: demographic, geographic, dis© 2014 Bentham Science Publishers

2

Adolescent Psychiatry, 2014, Vol. 4, No. 2

Carew et al.

cussion types, social media engagement and communications tools. The collected data was used to inform the redesign of the team’s website teenmentalhealth.org, as well as to provide strategic direction on tools and approaches that the team should consider to better meet mental health knowledge translation needs of its target communities. As mental health information in most communities is still being sought out behind closed doors, it was important to find out how and what information people were getting, and what they were searching for.

Documents Act) laws of Canada and protected under FOIPOP (Freedom of Information & Protection of Privacy Act). Content that is password protected is not provided and identification information is blinded.

METHODS

The search began with the collection process, where occurrences of the keywords (lexicon) were identified. The collection process is the most automated aspect of this research process. The team opted to use a dedicated search engine, MediaSphere360, to avoid the bias inherent in consumer search engines like Google or Bing. These consumer engines are designed with algorithms to deliver localized content and advertising. In addition, MediaSphere360 crawls into social networks, forums and other social media services that consumer engines do not. This provides a greater volume of relevant data, increased access to relevant conversations and a broader sample size.

The search included all forms of digital media between 2006 and 2010: social media (e.g. Facebook, YouTube, etc.), blogs and microblogs (e.g. Twitter), websites (including comments posted on sites), as well as newsgroups (e.g. alt feeds), bulletin boards (e.g. 4Chan) and forums. DATA COLLECTION Members of the team conducted the study with technical assistance from a social insight firm. The content analysis approach, used a search engine developed by the social insight firm called MediaSphere360. MediaSphere360, the search engine used in this study, extends the search of the Internet beyond commonly used consumer search tools, such as Google or Yahoo. This approach provided a means to identify and quantify who was looking for teen mental health information online, where they were looking, what they were looking for and conversing about, as well as how they preferred the information to be presented. By harnessing the power of online information, the team was able to identify and map patterns in online conversations. Instead of asking someone what they thought, or asking them to speculate what they would do in a particular instance, this approach allows researchers to track what people actually do. Specifically, this search tool has the ability to identify, monitor and track conversations based on a series of keywords and phrases. The team compiled over 100 terms and phrases (noted in Appendix A) most commonly referred to in clinical practice and the media related to mental health including slang references and youth terminology (such as: teenage depression, bulimia, obesity, gorging, addictions, etc.). Using these terms, a systematic search and synthesis was conducted to examine how youth and adults were searching for and engaging in conversations about adolescent mental health online. The search engine, MediaSphere360, collects all relevant data and uses its artificial intelligence to provide preliminary analysis on sentiment by putting words, phrases and memes into context. This is followed by a content analysis to organize the materials and findings. Given the large volume of data on this topic, the scope of the search focused on a cross-section of Internet usage between January 1, 2006 to October 2010 for English speakers residing in Canada and the United States. Groups were tracked based on patterns of online usage. This ensures that all information collected was under PIPEDA (Personal Information Protection and Electronic

DATA ANALYSIS A summative content analysis approach was applied to organize the materials and findings. This allowed the team to start with identified words and content, and search for their use in contextually (Hsieh, Shannon, 2005).

Once the collection was completed, the raw text was then entered into the Artificial Intelligence Engine (AI Engine) which is the overlay used to provide analysis of the text data sets. The AI Engine, through algorithms provides analysis on sentiment, gender, geographical location, entity extraction and correlation. These results were provided to the assigned human analyst. In addition, the AI Engine culls links and online channels and provides a set of links to relevant content. These links were assigned to a human analyst who then validated the information and provided a layer of contextual analysis behind the information. The content is also validated for spam content, sockpuppeting (online identity use for the purpose of deception) and astroturfing (the practice of masking sponsors of a message). RESULTS Demographics The findings of our search turned up more than 150 GB of data, which translated into 956,304 web pages that were analyzed and included 14,570 youth ages 13 to 18 and 14,235 adults with a total of 28,805 users. As shown in (Fig. 1), females were most active in online discussion related to teen mental health, youth (female 63%, male 27%) and adult users (66% female; 26% male). Note: approximately 9% of the youth users and 7% of the adult sample did not identify their sex in any of their online profiles or activities. Geography Geographic region was tracked by Internet service provider region (ISP), which revealed that in Canada, the Atlantic region had the highest online activity with respect to teen mental health (doubling from 2006-2010), followed by Central (which had consistent interest over the same period), and Western (which had dramatic decline in interest from 20092010). These results are depicted in (Fig. 2).

Using Digital and Social Media Metrics

Adolescent Psychiatry, 2014, Vol. 4, No. 2

3

Fig. (1). Demographics.

Fig. (2). Activity by Region.

The United States saw a similar pattern of interest, with North Eastern states having the most interest, followed by Western (interest declined dramatically from 2009 to 2010 in CA, WA, MT, ID, NV), followed by South Western, then South Eastern (with consistent interest between 2006-2010) and lastly Central (which showed growing interest since 2006). Social Media Engagement and Communications Tools The results of the social and media analysis revealed that between 2006 (baseline) - 2010 there was a 200% increase in online activity regarding teen mental health. Additionally the results show that adults and teens were using both similar and different online media channels (e.g. Facebook, blogging), but in different ways, and for different purposes (see Fig. 3).

Adults and youth visit different areas of the web. Adults were found to be most active in forums and newsgroups where ‘discussions’ are held. Adults were using Facebook as their prime social network tool. Adults will post stories and comments on blogs, but teens tend not to. The most popular microblog with adults was Twitter, where they tended to use this tool to share links to blogs, website and media articles. Conversely, youth were found to discuss mental health in newsgroups and forums, but not on blogs or Facebook. Teens often used social networks to lament suicides or tell related stories, but not to actually discuss issues. Teens tended to microblog less frequently then adults, and the age range for most microblogs trended between 25–55 years old. Teens who used microblogs tended to share stories and links, but rarely news articles. There are a number of blogs by teens on mental health experiences, but few comments were

4

Adolescent Psychiatry, 2014, Vol. 4, No. 2

Carew et al.

Fig. (3). Social Media Channels by Rank *Second coloured bar notes increase to total found in 2010.

left on them. Comments that were left are generally vague or empathetic. The most common newsgroups/forums used by both groups included: GoLiveWire, CityData, iMom, Tribe.net and ChristianRecovery. Discussion Types For adults, depression was the most often searched for topic, followed by suicide, doctors (most pronounced in US), anxiety disorders, addictions ADHD, sexual guilt, OCD and bipolar. For teens, the term suicide was the most searched topic, followed by depression, anxiety, sexual guilt, addictions, obesity, eating disorders and OCD. However, as (Fig. 4) indicates, information that adults and teens searched for did not necessarily result in increased conversations in these areas. Adults were most likely to ignite (initiate) conversations online about depression, followed by: anxiety, doctors, suicide, treatments, OCD, alcohol, and eating disorders. For teens, depression was also the most ignited (initiated) topic area, followed by: anxiety, alcohol, suicide, ‘sexting’ and marijuana. Adults were often seeking information about the disorders, treatment options and searching for other parent experiences. Teens tended to be discussing problems, concerns, or addictions through the use of personal stories. Note, the teen conversations about addictions were not about the desire to use the substance (Fig. 5).

Fig. (4). Hot Topics.

Adults and teens were both less likely to discuss at length any issues directly in social networks or blogs, but elected to have deeper and lengthy discussion in chat rooms, forums and newsgroups. DISCUSSION AND CONCLUSIONS The approach taken in this study is one that tracks the actual use of digital and social media by both young people and adults. This study provides a mental health snapshot in time. Information gathered using this methodology is reflective of a topic that is quickly evolving. Most notably this research showcases that a traditional approach to sharing information through one method to all audiences is no longer relevant or meaningful. The large sample size (28,805) provides substantial comfort that the findings are relatively reflective of the population using digital and social media. Perspective garnered about the various online channels used, with specific insight into how young people and adults access mental health resources when online, have several implications for mental health knowledge translation. The results of this study can provide mental health teams with a broader understanding of the key mental health topics of interest to both teens and adults, and how needs vary between both audiences. Through identifying the modes and methods used by youth and adults to search out information pertaining to men-

Using Digital and Social Media Metrics

Adolescent Psychiatry, 2014, Vol. 4, No. 2

5

Fig. (5). Ignited Topics.

tal health, the research team of this project was able to accurately reflect these needs and provide a portal to evidencebased content related to mental health (www. teenmentalhealth.org). Following the study, the website was redesigned to incorporate research findings such as increasing information in topic areas identified as being important to youth and adults. Additional features included: pioneering animated videos, presentations on a variety of topics from sleep to school mental health, social media and blog integration, youth mental health training programs and a library filled with resources that met the various needs of the target populations.

ography in the communication of information. For instance, the audience pool is largely weighted towards women. Resources should be designed to include female examples and imagery to appeal to this audience as well as empower females to communicate this information to males in their lives.

The preliminary results saw unique visitors to the website increase 1231% in the first year of launch. This measurement compares the unique visitors from 2008 (preresearch informed design changes) to 2011 (one year post launch). In addition, nearly 5000 resources were downloaded from the teenmentalhealth.org website in the first twelve months. These resources could be tracked directly back to the specific target audiences through on site self identification tools. The preliminary use of the data collected to create a resource portal demonstrated the impact of creating tools specific to identified target audiences’ needs, instead of creating tools that an institution wanted its target audiences to desire.

The quality and quantity of information this study provides stands in stark contrast to other approaches that rely on traditional survey or focus groups to seek insights. This approach can aid organizations in better addressing topic areas that are of importance to consumers as well as identify topics to be proactively monitored. Tracking over time can assist in identifying trends that can then be reflected as organizational communications evolve to continue to better address consumer needs. These research findings can be used to inform health care communications with emphasis on web-based health promotion materials and websites.

Other recommendations that stemmed from this study and its application included: the need to integrate messages and materials into multiple mediums, having a continuous stream of new and relevant information to share with an engaged audience, reflecting language tone and word selection to meet the audience needs, and the continued need to monitor topics and trends in the geographic location that one provides services within. In addition, it is essential that mental health communication strategies incorporate social media and other digital media tools (newsgroups, microblogs, etc.) where an identified need has been presented. The ability to provide substantiated, relevant data in locations in which the population is looking for answers in, is critical to closing the communication gap. The data gathered in this research offer several areas of potential future research, such as the role of gender and ge-

Geographically, it is shown that those in that coastal regions of both Canada and the United States have more online traffic related to mental health. An interesting area of additional research would be to relate these findings to services and/or programs offered in these geographic areas that may increase the mental health awareness of the public.

The application of this approach to ot...