Virtual reality and robotics technologies have developed interactions with augmented and replicated bodies in digital space. By simultaneously moving multiple bodies in various locations in ample space, users can reach distant objects and interact with them by switching the bodies without having to make significant movements themselves. Previous researches has investigated the design interfaces for manipulation and switching in such interactions using multiple bodies based upon the user's perception of the objects. However, the mechanism by which a self-attribution arises from multiple objects reflecting self-motion is still unclear. To investigate this mechanism, we developed a dummy cursor experiment in which participants were presented with multiple independently moving dummy cursors in addition to the cursors they moved on the screen. The presented dummy cursors moved in different directions and were temporally synchronized with the participants' movements. Participants identified the cursor whose movement exactly matched the mouse they were manipulating from among the multiple dummy cursors displayed on the screen. To investigate under what conditions participants could identify the cursor they were operating, we compared the identification rates by varying the number of dummy cursors presented and how they were moved. As a result, participants could identify the cursor they were operating with a very high probability, even when there were many dummy cursors. Through these results, we discuss the occurrence of a self-attribution among multiple objects related to self-motion and the possibility of parameters contributing to self-attribution.