Exploring Elderly Users’ Preferences for Interaction Design Features of Companion Robots for Daily Care in China

Abstract

Companion robots are increasingly being considered for supporting older adults in daily care, yet limited attention has been given to how specific interaction design features shape user preferences and willingness to use. This study examined interaction design preferences for companion robots in China using Self Determination Theory (SDT) as the guiding framework. A cross sectional mixed methods study was conducted with 120 survey respondents and five companion robot development professionals. Descriptive statistics, repeated measures analysis of variance, multiple regression, and thematic analysis were used to examine interaction preferences, design trade offs, predictors of willingness to use, and professional perspectives. Autonomy, competence, and relatedness support significantly predicted overall interaction preference, explaining 63.0% of the variance (R² = .630). Relatedness demonstrated the strongest association with overall preference (β = .479, p < .001) and was the only significant predictor of willingness to use (β = .314, p = .004). Respondents did not significantly differentiate among simplicity, warm interaction, and context appropriate proactive support. Expert perspectives converged with the quantitative findings, emphasising user control, simplicity, personalisation, social responsiveness, reliability, and family connectivity. The findings highlight the value of psychologically grounded and technically feasible interaction design that supports user autonomy and social connection while complementing rather than replacing human care relationships.