Here M is the number of spikes, Ui is the vector describing the x and y component of the spike-LFP phase for the ith. We only included units with at least 50 spiking events in the tested conditions to obtain a reliable and robust results and increase the statistical power of the analysis as has been shown before1569. We performed a paired t-test and also a permutation test to probe the significant difference of the PPC across experimental conditions. We have described the permutation procedure above. To correct for multiple comparisons, we have implemented the false discovery rate algorithm as described by Benjamini and Yekutieli70.
