- Introduction/Background The number of seniors in the world is growing and it is expected that there will be a growing need for home and nursing care services For this purpose, robots are expected to perform tasks of providing personal assistance and social care for the elderly Socially assistive robots (SAR) are a natural category for this endeavor In order for robots to be effective assistive tools, they must be accepted by the intended user population. Aims To test how age, location of touch interaction, and embodied presence of a humanoid robot affect the preferences of different age-group users when performing a cognitive-motor task Material and method A total of 60 subjects participated in two experiments The participants played a memory game of reach to grasp task using real physical cups with an SAR In the first experiment, 10 young (24.6 ± 2.6) and 10 older adults (70.8 ± 5.7) played 12 trials of the game with the robot In the second experiment, 20 young (24.75 ± 2.2) and 20 older adults (69.4 ± 5.64) randomly played 10 trials with the robot and 10 trials with a computer screen When finished, they completed a questionnaire. Results In both experiments participants preferred real human-like interaction with an SAR (In experiment 1: 70% P = 0.115; In experiment 2: P = 0.008 for petting the hand vs touching the tablet, P = 0.0022 in the older group), as they found it more “human-like” and “intimate” We found a significant preference for interacting with a robot over a screen ( P P P = 0.157 in the older group) Participants mentioned the robot was more “engaging”, “interesting” and “human-like”. Conclusion Both age groups preferred the interaction with a robot over a screen and specifically, real human-like interaction with a robot. The preferences of different populations, i.e. stroke, will be investigated in a future study.