Arranging 3D objects in Virtual Environments can be a complex, error prone and time consuming task, especially for users who are not familiar with interfaces for 3D navigation and object manipulation. In this paper, we analyze and compare novice users' performance on 3D object arrangement tasks using three interfaces that differ in the views of the 3D environment they provide: the first one is based only on a first-person view; the second one combines the first-person view and a map view in which the zoom level is manually controlled by the user; the third one extends the second with automated assistance in controlling the map zoom level during object manipulation. Our study shows that users without prior experience in 3D object arrangement prefer and actually benefit from having a map view in addition to a first person view in object arrangement tasks. Keywords 3D manipulation, virtual environments, user study, experimental evaluation.