De-identified Data for "AI-Assisted Autobiographical Storytelling with Older Adults: System Design and Epistemic Network Analysis of Narrative Meaning-Making"

This dataset accompanies the manuscript “AI-Assisted Autobiographical Storytelling with Older Adults: System Design and Epistemic Network Analysis of Narrative Meaning-Making.” The study used a six-week one-group pretest–posttest design with 60 older adults and examined changes in autobiographical narrative structure using Epistemic Network Analysis (ENA) and semantic alignment using BERTScore. The repository contains de-identified participant-level pre- and post-test data for six epistemic-code measures—Caregiver’s Emotional Journey (CM), Family and Interpersonal Relationships / Social Interaction (SI), Learning and Upbringing (LU), Childcare and Improvisational Creation (CI), Environmental Use and Adaptation (UE), and Conversational Fillers (TK)—together with ENA network density and BERTScore values. It also includes a codebook, supplementary methodological materials, and an analysis script for reproducing participant-level descriptive statistics and paired statistical comparisons. Raw autobiographical transcripts, audio recordings, direct identifiers, and other potentially identifiable narrative materials are not included because of privacy and ethical considerations.