Abstract: Fathers of Color's Preferences for Visualization and Features of an AI Chatbot to Support Paternal Mental Health Outcomes (Society for Social Work and Research 31st Annual Conference)

Schedule:

Saturday, January 23, 2027

Golden Gate 4 (Hilton San Francisco Union Square)

* noted as presenting author

Bryan Shannon, PhD, Research Associate, University of Maryland at College Park, College Park, MD

Hector Alcala, PhD, Assistant Professor, University of Maryland at College Park, College Park, MD

Craig Fryer, DrPH, Associate Professor, University of Maryland at College Park, College Park, MD

Elizabeth Norell, PhD, MSW, LCSW-C, Associate Professor, University of Maryland at College Park, College Park, MD

Mia Smith-Bynum, PhD, Professor and Department Chair, University of Maryland at College Park, College Park, MD

Background and Purpose: Fathers often report lacking support and resources during their transition into and through parenthood, with services more frequently tailored to mothers. Men from racial and ethnic minoritized backgrounds may also lack culturally relevant parenting resources. Artificial intelligence (AI) has recently been adopted to increase accessibility of parenting support for fathers; however, many existing resources do not reflect the needs of fathers of color. This study examines the content needs and visual preferences for an AI chatbot designed to support racial and ethnic minoritized fathers.

Methods: Utilizing a reflexive thematic approach, semi-structured interviews and focus groups (N = 4) were conducted with racial and ethnic minoritized fathers recruited via Prolific research participation system. We further explored racial and ethnic minoritized fathers' use of technology as a resource to support them as parents, as well as their content delivery needs and visual preferences of an AI chatbot to support them during their transition into and through fatherhood. Focus groups were audio-recorded and transcribed verbatim utilizing Rev human transcription and reviewed by members of the research team prior to analysis. Data were analyzed in NVivo utilizing a reflexive thematic analytic approach.

Results: Our reflexive thematic analysis identified three central themes: (1) what fathers want in “Rosie,” (2) visual and feature preferences, and (3) barriers and facilitators to using Rosie. More specifically, racial and ethnic minoritized fathers suggested the need for the inclusion of culturally relevant and vetted information within an AI chatbot to support them as fathers of color. Participants also stated the need for representation within the resource and among the development team of the resource, along with interactive community features such as a community discussion forum for fathers to support one another. Finally, fathers expressed that the quality of information within the resource, cost, and privacy served as both barriers and facilitators in adopting an AI chatbot to support fathers of color.

Conclusion: These findings highlight the specific needs for digital health parenting resources to support racial and ethnic minoritized fathers, who are often underrepresented and underserved. Racial and ethnic minoritized fathers seek digital parenting resources that provide culturally relevant information and representation. Findings suggest that future AI-based programming should incorporate culturally responsive content and interactive community features to enhance engagement and accessibility for fathers of color. Additionally, Rosie the Chatbot for fathers of color could transform the landscape of digital mental health resources to improve accessibility for an underserved group and improve family health outcomes.