8th Workshop on Ecological Psychology, Verona, 26-29 juin 2004
TIME INTERVALS PRODUCTION IN TAPPING AND OSCILLATORY MOTION Delignières, D., Lemoine, L. & Torre, K. EA 2991, University Montpellier I, France
Gilden, Thornton and Mallon (1995) performed spectral analyses on series of tapping intervals, and evidenced in the high-frequency region of the log-log plot of power spectra linear positive slopes, indicative of differenced white noise in the series. These results were interpreted on the basis on the Wing and Kristofferson (1973)‘s model, which contains such differenced noise. This result comforts the idea that a central, event-based timing process governs tapping performance. The aim of the present experiment was to show that the processes controlling the motor production of time intervals could depend on the nature of the required movements. We hypothesized that when time intervals are produced by an oscillatory motion, another timing model could be conceived, exploiting the limit cycle dynamics of the effector (Schöner, 2002). In such model, intervals are not bounded by discrete events, as in the WingKristofferson model, and power spectra should not exhibit the signature of differenced noise. Method 20 participants were randomly assigned to two groups. The first group performed a synchronization-continuation tapping task. The second group performed a similar synchronization-continuation task, but time intervals were produced through the oscillations of a joystick. Series of 700 successive time intervals were recorded. We analyzed the slope of the log-log plot of the power spectra, in the high-frequency region. Results The averaged power spectra are displayed in Figure 1. In the high-frequency region, the mean slope was 0.351 (SD = 0.317) for the tapping task, and –0.232 (SD = 0.220) for the oscillatory task. There was a significant difference between the two groups (F(1,18) = 25.894; p
