M. M. Schladen
Research Associate Professor, Biomedical Engineering
While autoethnographic methods offer valuable insights into caregiving experiences, traditional approaches rely heavily on memory and subjective interpretation, creating trust deficits in care decision-making. Family caregivers face uncertainty about behavioral patterns, medication effects, and health trajectories due to limited data collection and pattern recognition capacity. This formative research developed AI-Supported Caregiving Autoethnography, integrating consumer-grade cameras with systematic narrative documentation and interactive large language model (LLM) analysis to enhance observational trustworthiness.
We implemented multi-modal data collection combining Blink security camera snapshots with structured narrative logging over six months of elder care documentation. A systematic cognitive tracking scale was integrated with emergent narrative documentation. Consumer technology captured objective behavioral data while preserving authentic caregiver voice. Interactive LLM analysis enabled real-time equipment troubleshooting, interim pattern identification, and collaborative customization of observational frameworks.
The methodology created audit trails for care decisions, enabling retrospective validation of caregiver intuition through objective evidence. Interactive LLM collaboration identified correlations between medication timing, sleep quality, and cognitive performance that informed care optimization. LLM interaction enhanced caregiver confidence through real-time brainstorming and trend analysis. The approach democratized both rigorous longitudinal observation and sophisticated analytical capabilities, making systematic behavioral tracking accessible to individual family caregivers. Challenges included consumer-grade equipment performance variability and sustained documentation discipline requirements.
AI-Supported Caregiving Autoethnography addresses trust deficits by creating validation loops between objective and subjective data sources. This methodology transforms accessible consumer technology and democratized LLM tools into research infrastructure, enabling evidence-based caregiving decisions while preserving authentic autoethnographic voice. The interactive nature provides ongoing analytical support that enhances caregiver confidence and enables adaptive methodology refinement across diverse caregiving contexts.
To be presented at the TQR 17th Annual Conference—The Qualitative Report. "A Matter of Trust: Dates march 23-26, 2026