Abstract
How can one determine the nature of timing control that exists among various distinct phases of movement in a repetitive activity? A statistical approach is suggested and is exemplified by reference to data on the timing of two phases of movement, arrival at and departure from the response plate, in repetitive finger tapping. Two contrasting models are presented, and the predictions for the covariation of the various intervals between the two movement phases are compared. Data on finger tapping support the model that assumes the initiation of each phase is centrally determined without reference to the time of occurrence of the immediately preceding phase.