The Effects of Social Presence of Dementia-Prevention Chatbot on Isolation and Interpersonal Resilience Among Older Adults Living Alone: Evidence on Three-Wave Experiment (2026)

Jiyoung Lee, Woosang Hwang, and Sunyoung Park

Cyberpsychology, Behavior, and Social Networking

Abstract

Dementia-prevention technologies have mostly focused on cognitive enhancement, often neglecting the social aspects of healthy aging. This study examines whether the perceived social presence of a dementia-prevention chatbot, developed through sustained use, can reduce isolation and strengthen interpersonal resilience through communication networks among older adults living alone in Korea. A three-wave experimental design (N = 110) was implemented over 8 weeks using a smartphone-based chatbot that engaged participants through conversational cognitive games, health routine diaries, and social interaction prompts. Results indicated that perceived social presence of the chatbot at the midpoint significantly predicted greater interpersonal resilience both cross-sectionally and longitudinally, whereas its effects on isolation were nonsignificant. The results indicate that perceived social presence of the chatbot does not alleviate isolation but extends to interpersonal resources by reinforcing communicative resilience, thereby functioning as an indirect social pathway through which agent-based interactions may contribute to the reactivation of interpersonal relationships.

Fig. 1. Representative images of Sammy-Talk tasks: Cognitive Training Games (A–F) and Healthy Routine Diary (G) .

Notes: The right-side messages indicate participants’ messages. The left-side messages indicate Sammy’ s (virtual agent)

messages. The fi gure presents representative dialogue examples originally from a conversation in Korean from each task. In

the calculation game (A) , Sammy prompts participants to solve arithmetic problems and provides corrective or encouraging

feedback (e.g., “ Great job! Would you like to try another one?” ). In the language game (B) , participants rearrange letters to

form words. In the concentration game (C) , they identify hidden words in a grid. In the executive function game (D) , they

make planning or selection decisions. In the memory game (E) , participants recall previously presented words. In the spatial

perception game (F) , they select shapes that match a target stimulus. In the healthy routine diary task (G) , participants report

daily activities (e.g., exercise), and Sammy provides encouraging feedback and rewards (e.g., completion stamps). Across

tasks, Sammy delivers instructions and evaluative feedback, while participants provide task responses.