In our previous research we studied precision of tempo doubling and halving. Subjects were asked to double/halve a speed of their finger tapping. It was found that the precision of doubling/halving of a tempo was function of the base tempo. The task was performed the most precisely in the zone of intermediate tempi with IOI from 500 to 600 ms. The two distinctive tendencies were found. In fast tempi the difference from expected values corresponding to 2:1 (1:2) ratio decreased with the increase of a rate of the tempo, while in slower tempi the trend was the opposite.

In the present research we focused our attention to the order effect. It seems that the order of presentation of base tempi in a session influences precision of tempo doubling/halving. When we used during a session broader extent of base tempi (with IOI from 600 ms to 1500 ms) presented in a random order, we found responses, which were in contrast to the above described rules. It might be that large differences between neighboring tempi, which could by a chance appear during random order of stimulus presentation, impaired the performance. These tempi interfere with each other in short-term memory.

In series of experiments we systematically studied the effect of the order of stimulus presentation on precision of tempo doubling and halving. Base tempi were presented during a session sequentially from the fastest tempo to the slowest one (ascending order) or from the slowest tempo to the fastest one (descending order). This form of order of presentation of base tempi minimized temporal differences between neighboring tempi. We obtained results, which were consistent with our previous findings. In addition, a different effect of ascending and descending order of base tempo presentation in tempo doubling condition was observed. When the ascending order of stimulus presentation was employed, the performance was slightly more precise. In contrast, when the base tempi in the session were presented in order when temporal differences between neighboring tempi were maximized, we did not find any clear effect of a base tempo.

The results show that memory interference between presented base tempi is the factor influencing precision of the tempo change. The paper will discuss these memory interference mechanisms in more details.