This paper deals with the evaluation of human biofeedback response in virtual reality and in direct view. The experiments have been performed with a new paradigm for the evaluation of human biofeedback during the telemanipulation performance of a touch task. The controlled motion of one finger is monitored with the surface EMG, while mechanical robotized hand finger follows the motion imposed by the human finger. The biofeedback is detected in a direct way, by the vision of the robotized finger action, and in an indirect way, with the support of three different types of interfaces. The neuromuscular activity presents different features and delays in the four cases: A measurement of the attention and participation in the man/machine interface is obtained, in a first series of experiments. The paradigm adopted in this research is the result of the integration of robotics and neurology.
