Limitations in the use and interpretation of continuous relative phase.

Abstract

Continuous relative phase (CRP), a variable used to quantify intersegmental coordination, is difficult to interpret if care is not taken regarding the assumptions and limitations of the measure. Specifically, CRP is often interpreted as a higher resolution form of discrete relative phase (DRP). DRP, however, yields information regarding the relative dispersion of events in oscillatory signals while CRP describes their relationship in a higher order phase-plane domain. In this paper we address issues surrounding the calculation of CRP and suggest a new interpretation based on the aforementioned methodological issues. Through the use of test signals, with known properties, it was found that the CRP information will be arbitrary if no normalization procedures are used to account for frequency differences in the component oscillators. In addition, signals with non-sinusoidal trajectories will produce patterns in CRP that are not equivalent to discrete relative phase (DRP) measures. The implications of these issues are discussed.