Skip to main content
. 2015 Jul 29;35(30):10866–10877. doi: 10.1523/JNEUROSCI.4895-14.2015

Table 2.

Exploration of the effect of high-pass filtering and windows size used to estimate the autoregressive models

Comparison Pearson's correlation coefficient Rank correlation coefficient
Filtered versus unfiltered (500 ms window) 0.999 0.971
500 ms window (filtered) versus 350 ms window (unfiltered) 0.935 0.847
500 ms window (filtered) versus 750 ms window (unfiltered) 0.981 0.874

Neither filtering nor changes in the window size between 350 and 750 ms fundamentally affect the conclusions.