“You have people here to help you, people like me.” A Qualitative Analysis of a Blogging Intervention for Adolescents and Young Adults with Depression or Anxiety

Abstract

Blogging in the lay community has been shown to be a popular means of expression for all ages exhibiting mental illness symptoms. With the recent rise of mental illness rates among adolescents, blogging in a space specifically designated to discuss mental health topics for adolescents could potentially be beneficial for this demographic. In order to reveal whether or not blogging has positive effects on adolescents and young adults, we created a moderated, anonymous eHealth intervention for those in this demographic experiencing depression and/or anxiety symptoms. This intervention, called Supporting Our Valued Adolescents (SOVA), allows a safe place for participants ages 14–26 (inclusive) to read, write, and comment on blog posts regarding various mental health topics. In this paper, we analyze 40 SOVA blog posts and their corresponding comments written by 18 participants over a six-month period to see if actively engaging on the website was beneficial for their mental health. These posts and comments were analyzed on their degree of self-disclosure, regulatory and interpersonal support, acknowledgement of others, and reader feedback. We found that the content analyzed implied that blogging had a positive effect on participants using this online intervention.

Introduction

Background

Ninety-six percent of adolescents in the U.S. ages 13–17 consider anxiety and depression to be a problem among peers in their community, more distressing than bullying, drug addiction, and drinking alcohol (Horowitz, 2019). Despite the salience of this topic to youth and the potential for depression to lead to poor physical health, academic, and social outcomes, (Katon et al., 2010), most adolescents do not receive professional mental health treatment (SAMSHA, Center for Behavioral Health Statistics and Quality, National Survey on Drug Use and Health, 2016 and 2017., n.d.). Adolescents tend to try non-evidence-based treatments first as a form of self-help and self-reliance (Jennings et al., 2015). This is made even more feasible with the strong presence of technology in the lives of today’s youth (Townsend, Gearing, & Polyanskaya, 2012; Hickie, Luscombe, Davenport, Burns, & Highet, 2007), with 45% of adolescents aged 13–17 feeling that they are online on a near-constant basis (Anderson & Jiang, 2018). Given the rising rates of mental health diagnoses and the concern among adolescents about the state of their generation’s mental health, many are likely to use the Internet to look up information on psychopathology (Horgan, McCarthy, & Sweeney, 2013). Potential resources include professional articles, medical journals, and social media. More specifically, social media sites allow users to not only acquire new information, but also share their thoughts and concepts with one another while interacting with individuals from all over the world. Online communities through mediums such as blogging websites may cultivate social support, reduce feelings of loneliness, and allow opportunities to learn techniques to better handle adversity (Powell, McCarthy, & Eysenbach, 2003).

Expressive writing (defined here as detailed, personal writing that focuses on how the writer is feeling or has felt about events, relationships, and memories) can serve as an effective tool to benefit those experiencing severe mental illness (Pennebaker & Seagal, 1999). The most popular link connecting mental health recovery and expressive writing is through narrative therapy, where writing gives individuals the opportunity to take control of their own story and the power to define themselves instead of their mental illness defining them (King, Neilsen, & White, 2013). Expressive writing has also been found to reduce depressive symptoms, particularly before an impending, stressful event (Lepore, 1997).

Utilizing expressive writing on a social media site, such as writing blog posts with an interactive component, may be even more beneficial for several reasons. Firstly, because users write for an online audience, this offers an additional opportunity to provide and receive social support (Fiddian-Green, Kim, Gubrium, Larkey, & Peterson, 2019; Boniel-Nissim & Barak, 2013). Additionally, users may benefit from self-disclosure to others. Self-disclosure, or the extent of what one reveals about themselves, can enhance intimacy and well-being in the context of positive social support, and has been found to enhance subjective well-being and perceptions of social integration through blogging (Ko & Kuo, 2009). Self-disclosure is likely to occur in online support spaces and may promote positive mental health (Moreno et al, 2011) by inducing reciprocity of self-disclosure in others (Barak & Gluck-Ofri, 2007). Thirdly, the process of blogging for a supportive online audience may lead to potential psychological benefit through building strengths-based assets. The Resilience Portfolio Model (Grych, Hamby, & Banyard, 2015), describes how individuals may be able to use resources and strengths-based assets to lead to more positive appraisals of new situations and enhanced coping behaviors which in turn lead to improved psychological health. Blogging may promote building strengths-based assets such as regulatory, interpersonal, and meaning-making strengths. For example, through writing and sharing an experience, a blogger can demonstrate how they maintain positive or neutral affect even through a difficult situation by regulating one’s emotions and behavior. Building these assets can help serve as a protective factor from worsening mental illness symptoms (Zimmerman, 2013).

While blogging communities for individuals with mental health disorders are available, it is difficult to study the potential utility of blogging in a natural environment because researchers do not have access to user information and therefore cannot easily survey them to determine how blogging affects them, or access their content if it is private. Prior research in groups of adolescents have experimentally evaluated the potential benefit of blogging for social and emotional wellness, but these studies did not specifically target a population with depression or anxiety (Boniel-Nissim & Barak, 2013). We had the opportunity to learn more about whether potential benefits of blogging expand to specific groups of adolescents and young adults with a history of depression or anxiety, using an online research intervention we had developed for the purposes of increasing uptake of mental health treatment in adolescents referred by their primary care provider. Supporting Our Valued Adolescents (SOVA) is a stakeholder-informed (Radovic, DeMand, Gmelin, Stein, & Miller, 2018), technology-based intervention (Radovic, Gmelin, et al., 2018) with separate but related websites for adolescents (sova.pitt.edu) and parents (wisesova.pitt.edu) that targets: (1) decreasing negative health beliefs and psychoeducation through daily blog posts, (2) increasing social support through anonymous and moderated interactions with other adolescent users (Windler, Clair, Long, Boyle, & Radovic, 2019), and (3) enhancing parent-adolescent communication about mental health through parallel blog posts for parents and suggestions for conversation topics with their child. To encourage adolescent engagement with the material on the site and each other and to enhance the authenticity of the blog content, we added an opportunity for young people with a history of symptoms of depression or anxiety and interest in anonymously writing about their experiences to compose posts for and regularly comment on the sites.

This redesign of the SOVA intervention provided us with an opportunity to study more about blogging and its potential benefits. Specifically we wanted to learn: what topics will participants write about; to what degree will they self-disclose information or emotions; whether they will display aspects of the Resilience Portfolio Model including building interpersonal, regulatory, or meaning-making strengths; and to what extent they will recognize the existence of an online community and provide each other with social support in the context of a moderated online blogging intervention for adolescents with depression or anxiety.

Method

Procedure

Recruitment and Enrollment

Participants were recruited through advertisements on local university campuses, counselor referrals at those universities, local mental health clinicians seeing adolescent or young adult patients for psychotherapy, and the online University of Pittsburgh Research Registry, Pitt Plus Me. Counselors referring participants to the study did not directly send information to the research team but only provided a research flyer and did not offer the intervention in lieu of therapy. Those interested in becoming a participant and writing posts for the website were instructed to contact the study research team via email. The research team would respond to provide a brief summary about the purpose of the SOVA websites, blogging procedures, and compensation. After confirming that they wanted to continue, participants were sent the eligibility survey through Qualtrics (www.qualtrics.com). Participants were included in the study if they were between the ages of 14 and 26, inclusive; capable of reading and writing in English; completed the sixth grade; capable of accessing the Internet and email; self-reported current or prior symptoms of anxiety and/or depression; and were not deemed to be at high-risk for suicide. At screening, potential participants were asked if they have current suicidal thoughts with an intent to act and/or a prior suicide attempt with no subsequent follow-up treatment. All participants were provided with access to a list of crisis resources. Of note, no participant endorsed suicidality at eligibility during the time period studied.

If participants met eligibility criteria, they were then redirected to complete an online informed consent. A waiver of parental permission was approved by the IRB due to minimal risk and because we wanted to include participation from adolescents who may have poor communication quality with a parent and may have difficulty obtaining parental permission. Participants who completed consent were then asked to register for the SOVA website with an anonymous username. In the initial phase of recruitment, we did not collect demographic information to foster the feeling of anonymous contribution to the site and to protect the privacy of the participants.

Onboarding Process

After study entry, the SOVA advisor, a member of the research team with a background in mental health and acting as the main point of contact with participants, would contact each participant to explain the blogging process in greater detail either online, in person, or over a phone call. This interaction also served as an opportunity for the participants to ask any questions about the study. The advisor also served as the editor of the website, composing articles not written by users and therefore setting a standard for article style and content. Additionally, the advisor served as one of the website moderators, a role shared with other research members, who would screen new comments for safety and reply to comments with supportive messages if they were not replied to by another peer within 24 hours (Windler et al., 2019). During the advisor’s initial call with the participant, the advisor confirmed the participant understood the study is not a substitute for therapy. The advisor specifically reviewed potential topics the participant may be interested in writing about as well as tips about online health literacy (http://sova.pitt.edu/online-writing-style-guide). The advisor conveyed that the purpose of SOVA is to foster a supportive online community and so encouraged commenting on other users’ posts. The role of the advisor was shared by a research assistant with a master’s in social work and a graduate student in developmental psychology.

Participation included two components on a monthly basis. First, participants were encouraged to write at least one blog post per month on any topic they wanted pertaining to mental health, which were then classified under one of four categories: “Be Positive” (posted on Mondays), “Educate Yourself” (Tuesdays and Wednesdays), “Social Media Guide” (Thursdays), and “Links” (Fridays). Participants were asked to write original content only. They could email their post as an attached document or directly submit it via a website form on the SOVA website, hosted on WordPress. Each post was reviewed by the advisor prior to being published to the public website for accuracy, sensitive content, grammar, and provided feedback, if necessary, via email. Posts were run daily from Monday through Friday, except for major holidays, along with posts written by the research team. The second component was asking participants to leave at least four comments on other blog posts each month. These comments – which would post at the end of a blog article - could be in response to blogs written by other participants, or comments left on educational posts written by the research team.

Compensation was provided through prepaid debit cards, which were reloaded each month based on participation. Participants were paid $10 a month if they completed at least one blog post and an additional $5 if they submitted at least four comments that same month. There was no official duration for participation, although if participants turned age 27, they were asked to discontinue their registration with the site due to aging out.

Study Flow

See Figure 1 for more detail. During the almost 20-week time period studied, there were 44 adolescents and young adults who showed interest in becoming a participant. Out of these 28 completed the screening/eligibility survey and all met eligibility criteria and completed consent. These potential participants were sent an introductory email and asked to schedule an onboarding call, with 18 completing the call and thereby participating in the intervention.

Fig. 1.

A flowchart describing the recruitment process and study procedures

Analyses

Text from 40 blog posts, written by 13 participants, and their corresponding 183 comments, written by 18 participants (five of whom had not yet written any posts), were extracted from the SOVA website between September 25, 2017 through February 5, 2018 and entered into NVivo 11 qualitative software (QSR International Pty Ltd., 2015). The research team developed a prespecified, hierarchical codebook based on a desire to describe topics the participants chose to write about; whether participants exhibited self-disclosure and to what degree and what type of online self-disclosure based on descriptions by Barak and Gluck-Ofri (Barak & Gluck-Ofri, 2007); whether participants wrote about or exhibited strengths-based assets including regulatory, interpersonal, and meaning-making strengths as described by the Resilience Portfolio Model (Grych, Hamby, & Banyard, 2015); as well as evidence of participant online interaction. These initial code categories included: Blog Topic; Degree of Self-Disclosure (classifying the magnitude of expressed thoughts, feelings, and information); Regulatory Strengths (including building self-esteem and emotional self-efficacy); Interpersonal Strengths (including providing and seeking informational and emotional support, as well as feeling reductions in loneliness and stigma); Meaning Making (someone writing about what they learned from an experience); Acknowledging Others (users displaying language that indicated that they recognized the presence of the online community); and Reader Feedback (when another user provided feedback about the content they read, such as what they learned from it, that they felt motivated, or that they agreed with the content). The full codebook can be found in the Appendix.

A template analysis approach (King, 2004) and content analysis was used to develop a comprehensive summary of phenomena regarding main themes, while staying close to the data. The research team and stakeholder advisors using a team-based iterative approach met at the start, mid-way, and end of analyses to adapt the codebook, review extracted text corresponding to specific codes, and discuss emerging themes. Specifically, after the first 15 blogposts were coded, the full research team and members of our stakeholder advisory board with experience in mental health advocacy and blogging met to discuss discrepancies in codes and further revise the codebook. For example, the lowest degree of self-disclosure was eliminated from the codebook due to its inconsequential nature and lack of contribution to understanding self-disclosure. After the initial codebook was agreed on, two research assistants separately double-coded all blog posts and comments. These research assistants received training from the principal investigator. Individual codes that did not have 90% agreement were discussed by the research assistants with arbitration by the principal investigator until consensus was achieved; subsequently all posts were recoded until all codes achieved 90%. The research team and stakeholder advisors then met for a third time once all blog posts and comments were coded to review the codes and discuss the emerging themes. Approval for this study was received from the University of Pittsburgh IRB.

Results

Blog Topic

Each blog post was perused and classified under a topic based on the title and content. In some cases, several blog posts fell into more than one category and were double-coded. The majority of the 40 blog posts written over the time period analyzed were psychoeducational and/or shared personal stories, both with the purpose of providing information on a mental health topic and helping others (Table 1). These ranged from the participant’s own history of mental illness and seeking treatment to providing research, tips, and suggestions on how to manage mental health. Other topics included reviews of books, apps, and other media, and advice about navigating difficulties which may occur while using social media.

Table 1.

Blog Topic Example Codes

Subcategory (Total Coded)	Example Blog Posts
	Title	Description
Education about Mental Illness and Ways to Improve Mental Health (26)	A background on starting/stopping antidepressants	The author of this blog post answers frequently asked questions on antidepressants based on their own experiences.
Share a Story (13)	My Journey with Multiple Mental Health Diagnoses	A writer discusses what it is like for them to have multiple diagnoses in terms of treatment and stigma.
Review (App, Book) (4)	App Games for Anxiety	The author suggests some apps through a link and personal use that may be beneficial in relieving some anxiety.
Resource Sharing (4)	Journaling for Growth	The writer provides a link that discusses the health benefits that journaling can have.
Social Media Advice (2)	Why I Quit Social Media	The post reveals the writer’s reasons for deleting their social media and what happened as a result.
Artistic (Photo or Poem) (1)	Downpour	This ambassador shares how poetry is therapeutic for them, and then proceeds to share a poem they wrote about being yourself.

Self-Disclosure

We found participants were willing to reveal information about themselves, their thoughts, and their feelings to varying degrees (Table 2). Though users were anonymous and told not to reveal any information that could directly identify them, participants were still willing to share some personal identifying information (“I am an 18-year-old gay male, currently living in the Pittsburgh area” (ID 2). They also shared basic experiences that their peers may relate to, such as school activities or what hobbies they were interested in.

Table 2.

Degree of Self-Disclosure Example Codes

Subcategory	Brief Definition	Examples
Information 2	General information about the writer (age, occupation, interests, hobbies, etc.)	“Personally, one of my favorite things to do when I’m feeling more anxious than usual is to engage in game-play. I’m a huge fan of video games and have since come to enjoy downloading and playing different games on my smartphone.” (ID 6)
Information 3	More revealing information about the writer (traits, personal characteristics, problematic behaviors, etc.)	“My depression manifested itself in making me super irritable at times. Once I became aware of it, I had to constantly check myself and see if things that were irritating me were actually valid, or if I was just unreasonably moody.” (ID 7)
Thoughts 2	Expressing the writer’s personal thoughts on past or future events	“I believe this is the greatest message the show [Dear Evan Hansen] offers: we are not alone in our struggles.” (ID 3)
Thoughts 3	Expressing thoughts regarding the writer’s personal traits, appearance, ideas, etc.	“‘Like, really, another diagnosis?’ I thought. ‘This one isn’t even an anxiety disorder!’ I began to wonder if I was the only passenger on this multiple diagnosis trip!” (ID 16)
Feelings 2	Expressing mild feelings or ordinary concerns	“I was also surprised to find out that a lot of health care professors are not aware how to properly help someone going through a mental health crisis. I love seeing news like this; it is awesome to see this important development [becoming a mental health first-responder] in mental health research.” (ID 10)
Feelings 3	Expressing deep feelings, both positive and negative	“I know what it feels like to wake up and look forward to going back to bed again. I know what it’s like to feel like you are just going through the motions, but not really living. I have looked in the mirror and cried because I did not know the girl in the mirror. While most people may not have recognized that I was not okay by looking at me, I was able to visibly see the emptiness.” (ID 11)

Participants were willing to share their thoughts and feelings at a moderately personal level. Participants commented on how the posts made them feel while also showing support and appreciation by thanking the author for writing their article and sharing what the post taught them. Participants frequently did this by starting their sentences in the first person (“I didn’t know that...” “I love these tips,” “I’m happy to hear,” “I really enjoyed,” etc.). A few participants were also willing to share deeper, more personal experiences about themselves and their stronger emotional responses, which were typically more negative. In one post, a participant shared what is it like to live with obsessive compulsive disorder (OCD): “I remember lying on the floor in the foot of space between the bed and wall, urgently whispering confessions to my mom...” (ID 15). Another participant replied to a post about anxiety and panic attacks, describing their own experiences and advising, “It was really scary and I definitely recommend seeking help if you or anyone close to you has experienced something similar” (ID 6).

Regulatory Strengths

Few blog posts or comments expressed content mapping on to regulatory strengths (Table 3). Those that did expressed ways to strengthen self-esteem by sharing tips and strategies that personally work for the authors revealing them. Another regulatory strength, emotional self-efficacy, was present when participants openly discussed their personal improvement and increased confidence in their abilities.

Table 3.

Regulatory/Interpersonal Strengths Example Codes

Regulatory Strengths
Subcategory	Examples
Building Self-Esteem	“There is a bracelet that I wear everyday that reads, ‘You are enough.’ Anytime I am feeling down or worry that I am going to fall back into the darkness, I just look down, read my bracelet, and breathe. You, reading this right now, are enough. Never ever forget that.” (ID 11)
Building Emotional Self-Efficacy	“I deserve better and I will not compromise my happiness for unhealthy friendships and relationships anymore.” (ID 14)
Interpersonal Strengths
Subcategory	Examples
Reduced Loneliness	“This article also let me know that I am not alone when it comes to anxiety around the holidays. Before I read this article, I thought everyone else was only happy during the holidays.” (ID 14)
Social Support – Providing Informational Support Specifically Coping	“I, myself find journaling to be extremely therapeutic. Writing my thoughts out on paper help me to sort through the restless chatter that goes on in my head. When I write things down and turn repetitive thoughts into poems I learn something new about myself each time; I learn about my own mental health and even come to better understand what makes me happy as an individual.” (ID 6)
Social Support – Providing Emotional Support	“It does get better in time. Hold on. Stay positive. Do not quit on yourself. You are worth it. You have people here to help you, people like me.” (ID 11)
Social Support – Providing Informational Support	“Since our society focuses so much on physical abuse, oftentimes we don’t even know that emotional abuse exists or what is looks like. Here is a link to help people become aware of what emotional abuse looks like. And here is another. If you, or someone you know, is in an emotionally abusive relationship, please seek help, because the relationship might be taking a toll on your mental health.” (ID 14)
Social Support – Seeking Informational Support	“I’m having a hard time juggling the feelings of ‘this is my last semester of college, so I should have fun, go to dinner with friends, attend events, etc.’ and ‘I would like personal down time to relax.’ I don’t want to look back and feel like college was all work and no play, but I also want to take care of myself. Any advice?” (ID 13)

Interpersonal Strengths

As a social media entity, there were many opportunities for interpersonal interactions. There was a common theme for bloggers to mention that they were feeling less alone, often doing so explicitly (Table 3). Participants also reassured others that they were not alone in their adversities. The majority of the interpersonal strengths found were through social support between the bloggers. This was further divided into four categories: providing informational support, seeking informational support, providing emotional support, and seeking emotional support.

Bloggers provided informational support by using bulleted lists, telling personal stories, providing resources (e.g. TED talks, articles, apps, advice from their therapists/other professionals), and sharing creative outlets such as poetry. The informational support provided by the authors was based on their own experiences and usage. Participants often used blog posts to discuss how activities such as journaling, watching television, and arts and crafts helped them feel better. Other information that participants shared included a range of experiences: from taking antidepressants and participating in the Mental Health First Aid program, to musicals that centered around mental health and how they had impacted their lives.

Bloggers also sought informational support from their peers by explicitly asking for ideas and advice. This came from both comments in response to blog posts and the questions asked at the end of blog posts to incite conversation from others. Participants provided emotional support to their peers in both their posts and in response to other posts. Users were both implicit and explicit when providing each other with emotional support. One participant ended their post about being a member of the LGBTQ community by saying, “People might not agree, but who cares anyway; you’re too cool for them” (ID 2). Other users were more direct and provided reassurance that their struggles were only temporary and that they were not alone. There were no quotes pulled that represented seeking emotional support. No quotes represented bloggers’ reduced stigma.

Meaning Making

The blog posts and comments provided few to no signs of meaning making directly related to the blogging website and content. In a blog post reviewing the book Furiously Happy by Jenny Lawson, a participant realized that “I would never trade those things [not having a mental health disorder] for the empathy and kindness that I have gained from being knocked down countless times” (ID 13).

Acknowledging Others and Reader Feedback

Authors demonstrated that they were aware of the audience to whom they were speaking to: an online community of peers who were experiencing mental health concerns like depression and anxiety (Table 4). Authors would often directly acknowledge the audience primarily by using the second person.

Table 4.

Acknowledging Others - The Reader Audience Example Codes

Examples
“I am so extremely happy to hear that one of the articles helped you!” (ID 10)
“Also, maybe there’s someone else out there that is experiencing all these same things, and this will make you feel more comfortable to hear that it’s happening to me too!” (ID 8)
“I encourage anyone who feels this way to push through that, and write down all that you are feeling during the moment even if it feels funny typing it or putting it to paper. You really do feel a whole lot better once it is over.” (ID 10)
“I’m glad you also found that the medication helped with managing multiple symptoms, too, [ID 15]!” (ID 16)
“And I would like to share my personal experience for anyone else who might be considering stopping or starting medication.” (ID 10)

While not frequent, there were explicit instances where participants specifically mentioned, or “tagged,” the author in their comments. There were no instances of “tagging” in the blog posts. However, the tone and statements written in some posts made it evident that users knew they were part of an online community. Some instances were direct, as participants used statements in their posts and comments such as “To anyone reading this...” and “For those of you out there...”. Other users, however, took a more implicit approach. For example, ID 10 wrote a post in which they shared their experience with starting and stopping a selective serotonin reuptake inhibitor (SSRI). While ID 10 did not overtly address the audience, they still appeared to recognize its presence by displaying their experience and knowledge with the hope of making someone else feel less alone or providing them with useful, helpful information.

Blogger feedback expressed their motivation and desire to use the knowledge gained from the post content. The participants made it known that blog post content taught them new information and skills both implicitly and explicitly. Users also described how some content inspired them to utilize new coping strategies. They would also acknowledge the author and content of the post through questions in the comments such as “Do you have any particular tips/things you’ve found particularly helpful?” (ID 10).

Others provided agreement with the topic or appreciation on what was discussed. Agreement feedback included supportive comments and instances where users identified with the author’s feelings: “I can totally relate to the avoidance and thought rituals” (ID 15). Appreciation came in the form of expressing that they learned something from the post “I wasn’t aware that...” (ID 13); gratitude and thanking each other; and encouragement “This poem is really inspirational” (ID 17).

Discussion

The present study was conducted to understand whether adolescents and young adults with symptoms of depression and/or anxiety participating in a moderated online intervention with encouragement to compose monthly blog posts and comment on each others’ posts displayed evidence of online behaviors which may lead to potential psychological benefit. We found that participants mainly wrote about mental health educational topics but through the lens of their own lived experience and were willing to anonymously disclose their thoughts, feelings, and emotions. They self-disclosed moderate amounts of information about themselves and shared emotional experiences, displayed multiple instances of social support, but few explicitly wrote about meaning making. No participant experienced any known harm and, importantly, no participant left a negative comment for other users despite the anonymity of the site.

Increasing evidence supports that peers with lived experience serve an important and unique function in providing support to mental health patients by providing hope through self-disclosure, role-modeling, and the use of empathy, trust, acceptance, and understanding (Ali, Farrer, Gulliver, & Griffiths, 2015; Davidson, Bellamy, Guy, & Miller, 2012; Ibrahim et al., 2020; J A Naslund, Aschbrenner, Marsch, & Bartels, 2016; John A Naslund, Grande, Aschbrenner, & Elwyn, 2014; Prescott, Hanley, & Ujhelyi, 2017). Our participants seemed to offer this unique approach by using their previous knowledge and lived experiences to share the more distinctive characteristics of mental health that may not be taught through typical psychoeducation. Participants appeared to write content in ways that came across as educational, advisory, and supportive of their peers. Users provided evidence they were consciously writing for an audience and some expressed that the purpose of them using their own stories and experiences was to help others.

This altruistic theme is common among community-supported blogs (Horter, Stringer, Venis, & du Cros, 2014). Altruistic interactions were also present between participants exchanging interpersonal support. These patterns reflect previous blogging research where users explicitly mentioned that they hoped their blog posts could be of support for those going through a similar experience (Silvén Hagström & Toft, 2019). Using their own story while also providing facts appeared to make educational posts more personal, perhaps to better connect with other users and aid in decision-making.

Participants appeared to self-disclose on a continuum; levels of self-disclosure and anonymity varied. Some were more willing to disclose information that could be considered a vulnerability in some communities and cultures such as their identified gender and sexual orientation. This information was sometimes shared even when it was not relevant to the post. These unprompted incidences of self-disclosure support findings from Barak and Gluck-Ofri (2007) that describe how individuals are more likely to self-disclose in a support group specifically for a shared characteristic versus a general discussion forum. Knowing that their audience all experiences mental illness symptoms may have influenced participants to feel more secure and comfortable with disclosing identifying items due to developing attachment and interpersonal bonds through a common identity (Kraut & Resnick, 2012).

It was hypothesized that participants would write about their experiences with mental health and thereby share through the narrative process what they learned and how they made meaning from those experiences. While meaning making seemed to be present in some blog posts and comments, it is unclear if this actively occurred during the blogging process, or if it had already occurred in the past and then was being shared on the blog. There are several possible explanations for why we did not capture much meaning making, which include the brief length of the intervention and the likelihood that other extrinsic therapeutic approaches would contribute more strongly to meaning making.

The convenience and anonymity of blogging have been shown to promote active participation from users who have debilitating, stigmatized disorders, as they find anonymously disclosing personal information online more comfortable than disclosing in person (Powell et al., 2003; Qian & Scott, 2007). Blogging communities specifically designed for groups addressing health needs have historically existed. For mental health topics, blogging may be especially advantageous for several reasons. Blogging is convenient and provides an option to stay anonymous, thus allowing those with stigmatized illnesses to comfortably discuss topics and questions regarding their health (Powell et al., 2003). Although there are many examples of online community-supported spaces for those who share similar experiences and hardships (Hagstrom & Toft, 2019; Tsai, Crawford, & Strong, 2018), there are few, if any, moderated, safe spaces specifically created for adolescents coping with mental illnesses and their symptoms. Safety is a valid concern: while adolescents experiencing depression or anxiety symptoms using publicly available social media websites may gain social support, they also may experience negative consequences such as cyberbullying (Radovic et al., 2017). There is consequently a paucity of research on these community spaces and the potential impacts they have on their users. Existing online spaces are heterogeneous in factors which may limit potential benefits of blogging such as a lack of accurate information, moderation, and potential for harm due to negative online interactions (Radovic et al., 2017; Rainie et al, 2017).

Our results provide implications for existing emotional support websites and future research. One aspect for successful emotional support in peer interventions is facilitating a mutual aid response. In a mutual aid response, individuals must be able to exchange their ideas, feelings, and resources without a power dynamic at play (Seebohm et al., 2013). In researcher-led online communities, a power dynamic may be apparent if the researcher too strongly conveys their expertise or is overtly involved in the community, for example composing blogposts with high scientific literacy. This may discourage users from self-disclosing or providing mental health education in their own words. Communities which promote mutual aid are appealing to individuals wishing to discuss sensitive topics because they act as a safe space, without the threat of power dynamics. Among the many roles that moderators held, providing feedback and enforcing a set of community norms without overtly trying to control the content of blog posts or conversation were the most important for creating and maintaining mutual aid response. As the number of participants grew, the website moderator’s presence lessened, contributing to fostering a mutual aid response.

Our findings support that similar non-research interventions which allow adolescent and young adult communities to come together and provide support to their peers over a shared source of stress, which are moderated by professionals, may be beneficial. These online communities may be able to facilitate the provision of psychoeducation from a personal perspective while also allowing various levels of self-disclosure, advice, and overall support to their audience. Although the presence of meaning making was ambiguous, it has implications for future research. Conducting more in-depth interviews to better clarify and determine meaning making would allow researchers to better grasp what the users are gathering from these websites.

Limitations

There are several limitations to this study. First, during the recruitment process, a diagnosis was not necessary to participate, so it is unknown to what extent participants who self-reported a history of symptoms of depression or anxiety had actually experienced clinical symptoms. This was done as we felt asking for this level of detail may hinder reluctant participants to self-disclose on the site. Also, we wanted to recruit participants with a variety of experiences with their mental health, including diagnostic uncertainty and initial treatment-seeking behaviors. Also, no demographic information was collected in order to preserve anonymity. However, this makes it difficult to generalize the findings to certain gender, age, racial, ethnic, and socioeconomic groups. Because we found that participants were willing to self-disclose this information, future research will include the collection of demographic information in order to strengthen the study’s external validity. Although participants displayed behaviors that may be beneficial to enhancing resilience, we were limited in these interpretations by only analyzing their written content contributed to the websites. In the future, we plan to publish a separate study on a quantitative analysis of self-reported pre- and 3-month post survey findings in a larger sample of blogging participants. No cyberbullying or negative interactions have happened thus far on the website. This does, however, reflect recruitment and selection bias, considering that participants most likely joined the site in search of mental health education and support rather than joining to be hurtful to others. This could also be due to the Hawthorne effect, as participants understood they were being observed as part of a research study and were anonymous to each other but not anonymous to the researchers.

Conclusion

The SOVA intervention gave us a unique opportunity to examine whether adolescents blogging about their experiences with mental illness in a moderated and anonymous online space appeared to demonstrate potential benefits. SOVA participants were willing to self-disclose personal stories about their experiences to provide psychoeducational benefit and social support to other users who subsequently provided acknowledgment of that benefit in response. Future research will determine whether participants experienced self-reported benefits. Sites such as SOVA, which offer opportunities for blogging and social connection, are capable of not only educating today’s youth on mental health and illness but also creating a safe space for adolescents and young adults to provide each other with a unique source of help through online social support and mutual aid.