Sensory feedback is a highly researched area for upper limb prosthetics, for which the stimulation is applied to either the upper arm or lower arm with minimal justification for the chosen site. In this study, we compare the recognition of three stimulation sites and their sensitivities to the stimulation applied to the upper and lower arm sites from a mechanotactile stimulation device. Our stimulation device is a crank-based mechanotactile feedback system, which applies a combination of a pressure applied in the normal direction and tangential direction to the skin at the stimulation sites on the lower arm and upper arm. The recognition rate for six grip patterns was found to be not statistically different between the upper and lower sites. Further, through Just Noticeable Difference (JND) measurements, there was no statistical difference between the sensitivities of the upper arm and lower arm at four stimulation sites. This study contributes to the literature from the point of view of identifying the upper arm as the alternative site of mechanotactile feedback for transradial prosthetic hand users in comparison to the lower arm, which is primarily used for EMG electrodes to identify the user's intention to control their robotic prosthetic hand.