JMIR Publications

ABSTRACT

Background:

Despite swift roll-out of policies concerning social media and adolescent mental health, prior studies on this topic have tended to exclude youth voices and over-rely on simple survey or smartphone data, limiting participatory research, mixed-method data synthesis, and meta-inference formation.

Objective:

To include youth voices and generate meta-inferences, we propose SocialsVoice, a participatory convergent mixed-methods randomized design and Photovoice adaptation using social media clips. We describe how SocialsVoice allows for Grounded Socials-Visualization that triangulates three data types—(1) quantitative survey statistics, (2) participatory qualitative insights, (3) and participant-donated social media content—to create an ecologically valid, integrated database of real-world social media experiences that can be employed for meta-inference generation.

Methods:

Guided by the Critical Youth Empowerment Model’s six principles for the conduct of youth participatory action research, youth participants (N=41) were randomly assigned to share social media clips depicting positive or negative mental health content, as defined by youth. In seven video-chat group sessions, participants discussed clips, authored captions, screened co-created summary videos, and completed pre-posttests. Participants discussed their clips using the SHOWED method, a series of six questions and an established tool for community-based discussion.

Results:

With respect to SocialsVoice activities, youth overwhelmingly reaffirmed how SocialsVoice offered meaningful engagement. Youth emphasized how the study provided a safe, unique space for having mental health conversations that was previously absent in the life, to even reflect on their own mental health and the scope of mental health issues and perspectives in their communities and on social media. Participants shared how social media scavenging for community-based discussion was “fun”, “cool” and “interesting”. Most youth found the SHOWED questions to be helpful for organizing their thoughts, provoking deeper thinking and conscience raising about the topic of safety and utility of social media for mental health. A few participants suggested ideas for improving SocialsVoice including: real-time survey results, more meeting sessions, and an algorithm reset activity. In the end, this new insight about mental health struggles shared online led participants to want to learn how to become better allies for people coping with mental illnesses, and as bystanders and users of social media. A printed/digital book summarizing the youth’s findings was designed for lay audiences and disseminated to all participants and key stakeholders, meeting the Critical Youth Empowerment Model’s closing principles of generating sociopolitical change and space for critical reflection.

Conclusions:

SocialsVoice is culturally acceptable, relevant, appropriate, feasible, and safe for youth populations including those experiencing stigmatization. SocialsVoice goes beyond data science and uses partnered research to co-create an ecologically valid, integrated mixed-methods database of real-world social media experiences to investigate mental health impacts.