Figure 7.

Memory performance as a function of phase entrainment. A, Phase entrainment measure was averaged between 1 and 2 s for each trial. These single-trial phase entrainment values are plotted as a function of subsequent memory (remembered, purple; forgotten, green) in each phase offset condition. Error bars represent 95% confidence intervals of means in 0 remembered, 0 forgotten, 180 remembered, 180 forgotten conditions. B, Normalized slopes of logistic regression models for each phase offset condition were averaged between conditions for each subject. The logistic regression model was fitted to each subject's memory performance, against the single-trial phase entrainment measure (averaged between 1 and 2 s). The violin plot represents the distribution of the averaged normalized slope values. The black horizontal bars represent the mean of the data. Individual scatter points represent the averaged normalized slope values for each participant. Violin plots were created with the MATLAB extension violin.m (Hoffmann, 2015). C, Recall accuracy of each phase offset bin that consisted of trials which mean phase direction of each trial was 270, 180, 90, and 0°, ±45° respectively. Left, An illustration of how the four phase offset bins were defined. Four bins were centered at 0, 90, 180, 270°, with a bin boundary of ±45°. Right, Recall accuracy was calculated in each bin. The x-axis represents the centered phase angle of each bin. For each trial, the mean phase direction between 1 and 2 s was computed. Then, each trial was sorted into one of the four bins depending on the mean direction (e.g., if the mean phase direction of a trial was a value between −45 and 45°, the trial would be sorted into the 0° bin). The proportion of remembered trials was calculated for 12 trials that were randomly selected from each bin for each iteration. The recall accuracy was averaged across 10 iterations. Error bars represent 95% confidence intervals of means in each bin condition.