Figure 3. A: Grand-average ERP waveforms (contralateral minus ipsilateral activity relative to the memorized display hemifield) time-locked to the onset of the memory display at parieto-occipital electrodes (PO3, PO4, PO7, PO8, O1, and O2) for the different object configurations. For illustration purposes, the presented waveforms were low-pass filtered at 12 Hz (24 dB/octave). Scalp distribution maps were comparable across all components (PPC, N1pc, N2pc, and CDA), we therefore chose to present the point in time at which the respective difference waves (between grouped and ungrouped configurations) reached their maximum. B: Mean amplitudes of the PPC, N1pc, N2pc, and CDA components as a function of Object Configuration. Error bars denote 95% (within-subject) confidence intervals. Significant differences revealed by pairwise comparisons are indicated by asterisks; * p< .05. C: Correlations. The scatterplots show the relationship between individuals’ behavioral performance in grouped (left panel) and partially grouped (right panel) configurations and their corresponding PPC amplitudes. Solid lines indicate the best-fitting regressions, shaded regions illustrate 95% confidence intervals.
Oscillatory Amplitude. Figure 4A shows the time–frequency profile across trials. As can be seen, variations of grouping strength modulated activations in the alpha band (8–12 Hz), with no other frequency ranges showing comparable changes in activity. Given this, we examined changes in the lateralized alpha amplitude (contralateral–ipsilateral) as a function of Object Configuration using a repeated-measures ANOVA with the within-subject factor Object Configuration (ungrouped, partially grouped, and grouped). The lateralized alpha during the pre-stimulus period (-200–0 ms) yielded no effect of configuration (-0.41 µV, -0.49 µV, and -0.37 µV for ungrouped, partially grouped, and grouped configurations, respectively),F (2, 46) = 0.60, p = .55,ηp2 = .03,BF10 = 0.18, showing that alpha amplitudes were comparable across conditions before trial onset. However, during thedelay period (350–1300 ms), there was a significant main effect of Configuration, F (2, 46) = 8.73, p < .001,ηp2 = .28, indicating that lateralized alpha is suppressed the most for ungrouped (-0.76 µV), followed by partially grouped (-0.69 µV), and least for grouped (-0.62 µV) configurations (all p’ s < .02, |dz |s > 0.44, for the pairwise comparisons between configurations; see Figures 4B and C ). Note that the observed differences in the lateralized alpha amplitudes were mainly associated with contralateral variations (which was strongly modulated by Object Configuration, F (2, 46) = 7.71,p = .001, ηp2 = .25), thus reflecting processing of the task-relevant stimulus configurations rather than the inhibition of task-irrelevant placeholders (the latter being associated primarily with ipsilateral alpha activity, which was overall comparable for different configurations, F (2, 46) = 0.12, p = .88, ηp2 = .005, BF10 = 0.13).