ABSTRACT

The behavior of complex systems exhibits recurrence. For music performance, recurrence has been used to examine timing and postural sway. This chapter describes how to measure recurrence in the movements of performers, but the same methods can be applied to the movements of listeners. A popular method for examining self-similarity in a complex system is recurrence quantification analysis (RQA). The chapter discusses two inferential approaches. The first reduces the RQA plot to a one-dimensional time-series and uses mixed-effect models to do the statistical analysis. The second keeps the data plotted in matrix form and directly compares plots using additional RQA techniques: cross RQA (CRQA) and joint (JRQA). CRQA and JRQA require that the performances compared be on the same timescale, which requires converting clock time to musical time, by time-warping, the performances. There are many methods of time-warping, including functional data analysis (FDA). The level of correspondence between performances provided by the C/JRQA plots can be assessed for statistical significance.