An experiment was conducted to analyze four different methods for combining object rotation and positioning in a two-dimensional plane. These methods examine the use of a prototype of a two-ball mouse, which senses not only the traditional x-axis and y-axis displacement, but also the z-axis angular motion. The latter is used as a possible improvement. Rotational sensing is accomplished through a simple calculation on the two sets of x-y displacement data. This allowed for rotation to be detected and performed using simple primitive operations. The added flexibility in a two-dimensional plane provides for much greater ease in accomplishing the tasks. The common rotate tool, found in most drawing programs, was compared with this three-degree-of-freedom mouse from the task perspective. The results showed a preference to using the three-degree-of-freedom methods both from a time (F\sb3,36=4.14, p<.05) (quantitative) and user (93% of participants chose three-degree-of-freedom methods) (qualitative) perspective. This paper discusses mechanisms to engage the added degree of freedom and different interaction techniques. Further comparative studies are performed to show the benefit and ease of use afforded to the end user.