1203 - MANAGING COGNITIVE LOAD WITH HUMAN AND ARTIFICIAL SUPPORT: AGE-RELATED DIFFERENCES ACROSS OBJECTIVE, SUBJECTIVE, AND PHYSIOLOGICAL MEASURES

Session: D07S002 - Cognitive Aging
AUTHORS:
Varrasi Simone (University of Catania ~ Catania ~ Italy) , Vagnetti Roberto (University of Manchester ~ Manchester ~ United Kingdom) , Camp Nicola (Nottingham Trent University ~ Nottingham ~ United Kingdom) , Hough John (Nottingham Trent University ~ Nottingham ~ United Kingdom) , Di Nuovo Alessandro (Sheffield Hallam University ~ Sheffield ~ United Kingdom) , Castellano Sabrina (University of Catania ~ Catania ~ Italy) , Magistro Daniele (University of Southampton ~ Southampton ~ United Kingdom)
Abstract text:
Introduction. Preserving autonomy in aging requires strategies to manage cognitive load (CL), a central determinant of performance and quality of life. Assistive technologies, including socially assistive robots (SAR), have been proposed to support older adults, yet it remains unclear whether they reduce or increase mental demands compared with human support.


Purpose. This study examined how human-human interaction (HHI) and human-robot interaction (HRI) influence CL during cognitive task performance, with a focus on age-related differences.


Method. Sixty healthy adults participated: 30 younger (M = 34.8 years) and 30 older (M = 72.3 years). Participants completed a cognitive task adapted from the Trail Making Test under seven conditions: baseline (independent completion), three HHI conditions (low, medium, high CL), and three HRI conditions (low, medium, high CL). CL was assessed via objective measures (accuracy, completion time), subjective ratings (NASA Task Load Index), and physiological indicators (salivary cortisol). Data were analyzed using mixed-effects models and ANOVAs with post hoc tests.


Results. Both HHI and HRI enhanced performance compared to independent completion. However, older adults showed reduced accuracy and slower performance than younger adults, particularly under medium and high CL. HHI consistently supported higher accuracy and lower subjective CL across groups. In contrast, HRI was associated with higher perceived strain and cortisol reactivity in older adults, suggesting additional cognitive demands. Younger adults exhibited no significant differences between HHI and HRI.


Conclusions. Findings highlight that while robotic support offers scalable assistance, current implementations may impose extraneous CL on older adults. Human support remains more effective in facilitating performance and reducing mental strain. Refining robotic interfaces to better align with age-related cognitive profiles is essential for developing adaptive, user-centered assistive technologies that promote active and sustainable aging.