Jeong Youn-Gill, Shin Seo Jung, Lee Song Yi
School of Interdisciplinary Studies, Dongguk University-Seoul, 30, Pildong-ro 1 gil, Jung-gu, Seoul 04620, Republic of Korea.
Department of Counselling and Coaching, Dongguk University-Seoul, 30, Pildong-ro 1 gil, Jung-gu, Seoul 04620, Republic of Korea.
Healthcare (Basel). 2025 Jun 26;13(13):1525. doi: 10.3390/healthcare13131525.
This study examines the subjective perceptions of older adults with mild chronic conditions regarding an artificial intelligence (AI)-based caregiving device, referred to as an "AI Human", by identifying and categorising their viewpoints through Q methodology. We conducted the study in February 2025 at two adult welfare centres in Buyeo, South Korea. Thirteen older adults used the AI Human device with support for at least 15 days. We initially generated 152 opinion statements through a literature review, focus group interviews, and AI-assisted methods and refined them to a Q sample of 34 statements. Participants completed Q sorts, and we used Ken-Q Analysis software (version 2.0.1) to analyse the data, applying principal component analysis and Varimax rotation. Four distinct perception types emerged: (1) emotionally engaged users prioritise reminiscence and emotional interaction; (2) present-oriented conversationalists prefer real-time, everyday dialogue; (3) usage-burdened users are interested in the device but experience usage difficulty; and (4) function-oriented users value health and caregiving functions. The acceptance of AI caregiving devices among older adults varies based on their emotional needs, conversation preferences, technical accessibility, and perceived usefulness. This study provides theoretical and practical insights for developing personalised, elder-friendly AI systems that support ageing and promote emotional well-being.
本研究通过Q方法识别和分类老年人的观点，考察患有轻度慢性病的老年人对一种名为“AI人类”的人工智能护理设备的主观认知。我们于2025年2月在韩国扶余的两个成人福利中心开展了这项研究。13名老年人在得到支持的情况下使用AI人类设备至少15天。我们最初通过文献综述、焦点小组访谈和人工智能辅助方法生成了152条意见陈述，并将其提炼为一个包含34条陈述的Q样本。参与者完成了Q分类，我们使用Ken-Q分析软件（2.0.1版）对数据进行分析，应用主成分分析和方差最大化旋转。出现了四种不同的认知类型：（1）情感投入型用户优先考虑回忆和情感互动；（2）关注当下的健谈者更喜欢实时的日常对话；（3）使用负担型用户对该设备感兴趣但在使用上有困难；（4）功能导向型用户重视健康和护理功能。老年人对人工智能护理设备的接受程度因其情感需求、对话偏好、技术可及性和感知有用性而异。本研究为开发支持老龄化和促进情感幸福的个性化、老年人友好型人工智能系统提供了理论和实践见解。
