Purpose:
The effects of practice were analyzed in the control of fast and accurate spatially constrained movements.

Methods:
Twenty men (20–26 years old) evenly divided into an experimental and control group were analyzed in three time periods: pre-test, post-test, and retention. Discrete Aiming Task ver. 2.0 software simulated Fitts’ task (1954) and provided kinematic analysis of mouse cursor movements (displacement, velocity, and acceleration). The task consisted of using the mouse to click on two parallel targets as fast and accurately as possible. Four target widths (W = 2.0, 1.0, 0.5, and 2.5 inches) and three distances between the targets (D = 2.0, 4.0, and 8.0 inches) were used to provide indexes of difficulty (ID) from 1 to 6 bits. The experimental group performed 108 practice trials (three blocks of 36 trials on different days) while the control group had no practice.

Results:
Movement time (MT) decreased in the experimental group largely due in part to a reduction of time used for feedback. It is suggested that the improvement in performance as a function of practice occurred through the interdependence of programming and the feedback process. As the task was practiced, there was decreased need for feedback due to better pre-programming of the primary submovement and the improved use of sensorial feedback information. This strategy and a lengthened deceleration phase can help explain the paradigm of fast and accurate movement as a result of practice.

Conclusions:
Despite the improved performance changes as a consequence of practice, Fitts’ Law proved to be robust enough to predict MT as a function of ID.