Effects of accuracy feedback on fractal characteristics of time estimation.

Nikita A Kuznetsov,Sebastian Wallot

doi:10.3389/fnint.2011.00062

Abstract

The current experiment investigated the effect of visual accuracy feedback on the structure of variability of time interval estimates in the continuation tapping paradigm. Participants were asked to repeatedly estimate a 1-s interval for a prolonged period of time by tapping their index finger. In some conditions, participants received accuracy feedback after every estimate, whereas in other conditions, no feedback was given. Also, the likelihood of receiving visual feedback was manipulated by adjusting the tolerance band around the 1-s target interval so that feedback was displayed only if the temporal estimate deviated from the target interval by more than 50, 100, or 200 ms respectively. We analyzed the structure of variability of the inter-tap intervals with fractal and multifractal methods that allow for a quantification of complex long-range correlation patterns in the timing performance. Our results indicate that feedback changes the long-range correlation structure of time estimates: Increased amounts of feedback lead to a decrease in fractal long-range correlations, as well to a decrease in the magnitude of local fluctuations in the performance. The multifractal characteristics of the time estimates were not impacted by the presence of accuracy feedback. Nevertheless, most of the data sets show significant multifractal signatures. We interpret these findings as showing that feedback acts to constrain and possibly reorganize timing performance. Implications for mechanistic and complex systems-based theories of timing behavior are discussed.

Highlights

Human performance on virtually all tasks exhibits variability from trial to trial (Bernstein, 1967)
MULTIFRACTAL SPECTRUM In terms of the multifractal spectrum analysis, our results show that time estimation performance is more multifractal than suggested by Ihlen and Vereijken (2010) when they reanalyzed the data from the Wagenmakers et al (2004) time estimation study
As Castillo et al (2011) put it: “The observed timing of physiology and behavior is an outcome of the coordination of the body, not the other way around.”. This experiment examined the effect of accuracy feedback on the long-term correlation structure of the time interval estimates in the continuation time estimation task

Summary

Introduction

Human performance on virtually all tasks exhibits variability from trial to trial (Bernstein, 1967). One of the first attempts to understand the nature of variability of discrete human movements was made by Wing and Kristofferson (1973) using tasks that required continuous estimation of a constant time interval by tapping the index finger. They hypothesized that two independent processes determine the inter-tap interval (ITI) at tap j (ITIj ) by the following relation: ITIj = Cj + Dj − Dj−1 , (1). Where Cj is a timing motor command from a central cognitive timer to the motor periphery that executes it with a neuromuscular delay (Di) compounded with the delay on the previous tap Both C and D were assumed to be independent, uncorrelated white noise processes. This prediction was verified in studies of Vorberg and Wing (1996), as well as Wing (1980) for series of about 100 estimates (cf. Delignières and Torre, 2011)

Objectives

Results

Discussion

Conclusion