Skip to main content
. 2021 Oct 14;3:753062. doi: 10.3389/fspor.2021.753062

Figure 5.

Figure 5

Illustration of the covariation by randomization (CR) analysis. All variables are randomly shuffled across trials to produce decorrelated data. This procedure allows for removing any correlation between variables within a trial. The decorrelated data are substituted into a forward dynamics model to determine the performance outcome and the standard deviation (SD) after the random shuffle (SDshuf). If SDshuf is higher than the variability of the performance outcome for the original data (SDactu), the coordination between variables is present to reduce the variability of the performance outcome.