Improving mental wellness often requires adopting new practices and thinking processes, which can be challenging due to a lack of motivation and difficulty visualizing how new strategies apply to personal lives. This research proposes that real-life stories, delivered via text messaging, can serve as an engaging method for promoting mental wellness by illustrating psychological principles in an applied context. Stories have historically been used to share knowledge and experience and can normalize challenges, validating the difficulty of making changes. Text messaging is chosen as a delivery medium due to its accessibility and widespread use. The study investigates three research questions: 1) What features of story-based text messages motivate people to apply mental health lessons? 2) How can story-based text messaging elicit self-reflection regarding mental health? 3) What design tensions arise when deploying such interventions? The study focuses on young adults (18-25 years old) due to their high prevalence of mental health concerns and high mobile phone usage. A formative study involving interviews and focus groups informs the design of a story-based SMS service, which is then deployed to 42 participants, followed by further interviews to understand user perception and benefits.

The literature review examines existing research on storytelling in mental health interventions and the use of text messaging for behavior change and mental wellness. It highlights the effectiveness of narrative-oriented approaches, such as entertainment-education and narrative persuasion, in influencing attitudes and behaviors related to various health issues. Existing evidence supports the efficacy of stories in reducing depression symptoms and promoting healthy habits. Digital tools for sharing peer narratives, such as online support groups and social media, are also discussed, emphasizing the role of peer interaction in behavior change and mental health management. The review also examines prior work on text messaging interventions for behavior change, focusing on its success in promoting abstinence from smoking and alcohol, as well as weight management and improved medication adherence. It includes discussion of chatbots designed to encourage self-reflection on mental wellbeing and the limited prior work using stories for mental health via text messaging. The final part covers cognitive distortions—exaggerated or unhelpful thought patterns linked to affective disorders—which are central to the study's story content.

The study employed a mixed-methods approach. A formative study was conducted first, comprising six semi-structured interviews and four focus group discussions with 15 participants (ages 18-25). Participants included undergraduate and graduate students in computer science and psychology, recruited through snowball sampling, and individuals with moderate to high levels of depression or anxiety symptoms, recruited through Mental Health America. Data collection involved questions on participants' experiences with mental health applications and ideas for story-based text messages, exploring desired features, such as authenticity and concrete takeaways. Thematic analysis was used to identify key themes. Following the formative study, the researchers designed a series of true stories centered on cognitive distortions (overgeneralization, all-or-nothing thinking, and fortune-telling), common in cognitive behavioral therapy (CBT). These stories were then deployed via text messages to 42 participants (ages 18-25), recruited through snowball sampling and Mental Health America. The deployment tested various features, including story length (short and long versions) and interactivity (Up-Front Choice and Start and Switch-Out formats for selecting stories). The Up-Front Choice format gave participants a brief summary to gauge interest before receiving the full story. The Start and Switch-Out format sent a portion of the story, followed by a question if they wanted to continue. After each story, participants received a relatability prompt and, if they responded positively or did not respond, a reflection prompt. Finally, 20 participants completed post-deployment interviews. Thematic analysis was employed for qualitative data from the formative study and post-deployment interviews. Quantitative data from the deployment study was analyzed through response rates to prompts and average word count in open-ended reflection prompts.

The formative study revealed several key themes: Participants valued authenticity in stories, preferring narratives from real people and concrete details. They also desired stories with clear takeaways, explicitly connecting the narrative to a psychological concept. Active participation, such as responding to reflection prompts, was seen as beneficial. Finally, participants sought a balance between realistic depictions of struggles and positive conclusions. The deployment study indicated moderate to high engagement with the story-based messages. Relatability prompts had a high response rate, and participants wrote detailed responses to reflection prompts, showing self-reflection and application of story lessons to their lives. The feedback largely corroborated the formative study findings: Participants found the stories authentic and appreciated the concrete takeaways, which helped them translate the lessons into their own experiences. However, there were varying preferences for story length, with some finding longer stories more detailed and engaging, while others found them overwhelming. Both the Up-Front Choice and Start and Switch-Out story selection formats had their respective pros and cons, with some users preferring choice and others wanting the option to continue or switch mid-story. However, abrupt endings in the Start and Switch-Out format were identified as a concern.

The findings address the research questions by demonstrating the efficacy of story-based text messages as a mental wellness intervention. The use of real-life stories, emphasizing authenticity and concrete takeaways, proved effective in engaging participants and promoting self-reflection. The study suggests that a balanced approach, combining realism with positive outcomes, is crucial for designing such interventions. The varying preferences for story length and format highlight the need for flexible and adaptable designs, potentially incorporating options for both short and long narratives and different methods of story selection. This study contributes to the CSCW literature by showing how low-stakes, accessible interactions with an automated system can maintain important benefits of sociality in mental wellness interventions. The findings indicate that even without direct peer interaction, learning about others' experiences can create a sense of connectedness and encourage self-reflection, which can lead to meaningful changes in thought patterns and behavior. The potential for future work includes improving the interface for longer stories through a dedicated app, enhancing personalization of stories, and integrating the stories with online resources.

This paper demonstrates the potential of story-based text messages for improving mental wellness. The findings highlight the importance of authenticity, concrete takeaways, active participation, and a balanced portrayal of struggles and positive outcomes in designing effective interventions. Future research should focus on improving interface design for longer narratives, enhancing personalization, integrating online resources, and exploring efficient methods for large-scale story generation, such as user-generated content and crowdsourcing. This approach offers a scalable and accessible way to support mental health.

The study's limitations include a focus on young adults in North America, potentially limiting the generalizability of findings. The relatively small sample size restricts the ability to draw strong conclusions about differences between participants with and without mental health symptoms. The research also focused on a limited subset of cognitive distortions and did not conduct an extended longitudinal deployment. Future research should address these limitations by recruiting a larger, more diverse sample and conducting longer-term deployments.