A robot goes to rehab: a novel gamified system for long-term stroke rehabilitation using a socially assistive robot-methodology and usability testing.

The study measured only how patients felt about the system. It did not test whether the platform improved arm or hand function, so it cannot tell you whether the exercises work. The sample is very small: 24 randomised, 18 completing, and the groups were unequal (11 robot, 13 computer). The authors say the lack of difference between groups may be because the study was too small to detect one. The robot group also received conventional therapy alongside the platform, so any benefit cannot be separated from their usual treatment. The two groups differed only in whether a robot or a screen delivered the instructions, and the robot version included welcoming phrases and victory gestures that the screen version did not, so the comparison is not a clean test of a robot versus a computer. Four participants were dropped because the unit closed for the COVID-19 lockdown, and two more in the computer group quit because they did not want to continue, which may have left a less representative group completing the study. The satisfaction questionnaire was custom-made and not a standardised measure, and the authors acknowledge they tried to balance positively and negatively worded questions. The colour-coded games were not tested with people who are colour-blind.

Funding came from the Israel Science Fund, Rosetrees Trust, the Borten Family Foundation, the Consolidated Anti-Aging Foundation, the Helmsley Charitable Trust through a robotics initiative at Ben-Gurion University, the Marcus Endowment Fund at Ben-Gurion University, the Israeli Institute of Social Security, the European Union's Horizon 2020 programme, the McCormick Foundation and the Robert Bergida bequest. The system was built by the authors' own lab team. No commercial conflict of interest is declared in the text supplied.

Do not read the higher robot ratings or the zero voluntary dropouts in the robot group as proof that a socially assistive robot motivates patients more than a computer screen. None of the satisfaction differences were statistically significant, and the paper states plainly that the lack of difference may be due to the small sample size. The authors' own conclusion is that motivation tracked whether patients believed the system helped them, not whether a robot was present.