To investigate θ power recorded at rest with EO, Fourier-transform (FT)-based spectral analysis1 was applied to the EO resting data. The resting data were first corrected for ocular artifacts (Gratton et al., 1983). The data were then divided into non-overlapping segments of 128 data points (corresponding to 512 ms and matching the segment length utilized in the computation of the STFT-based event-related TFRs for the oddball task). Any segments containing out of range voltages (± 75 μV) were rejected as artifacts. A Hanning windowed was applied to each segment and the segment was zero-padded with additional 384 data points (matching the windowing and zero-padding applied in the STFT-based event-related TFR computation for the oddball task), and the complex-valued Fourier transform of the zero-padded, windowed data in each segment was computed. The square of the magnitude of each single-segment complex-valued spectrum was then computed and averaged across all segments. Theta band power was computed from the resting power spectra by computing the average power within the same θ frequency band as used for the event-related power measures in the task.
