To test whether the means of paired observations were significantly different, we used a permutation paired t test, in which a paired-sample t-statistic was computed on the observed data, and on data with the labels of each pair of observations randomly swapped. To test significance of multiple main effects and their interaction, we used a permutation 2-way ANOVA in which an F-statistic was computed on the observed data, and on data where the multi-factor labels were randomly shuffled as a group across trials. All tests were corrected for multiple comparisons across time points and/or frequencies using a procedure that controls the false discovery rate under arbitrary dependence assumptions (Benjamini and Yekutieli, 2001) using the Python statsmodels module.
