Figures and data
Histogram (A) and bar graph (B) showing the distribution of ASRS scores in the AD(H)D and Control groups. Error bars represent standard error of the mean (SEM). This confirms group-allocation, within the tested sample.
A. Illustration of the Virtual Classroom used in the study. Inset: a participant wearing the VR headset over the EEG cap. B. Example of the sequence of sound-events presented in Event-trials, which were emitted from a spatial location to the right or left of the participant. Event-trials contained 5 events, of both types (Artificial and Human non-verbal sounds), randomized across trials and separated by 3-7 seconds. C. Example of a multiple-choice comprehension question presented after each trial (English translation).
Accuracy on comprehension questions.
Shown separately for the AD(H)D and Control groups, in the Quiet and Events conditions. Bar graphs represent average accuracy levels across participants and error bars represent the SEM. No significant differences were found between groups in either condition (n.s).
Speech-tracking of the teacher in the presence of sound-events.
A. Topographical distribution of the predictive power values, estimated using the multivariate encoding model, in the ADHD and Control groups. No significant differences were observed between groups. B. Grand-average TRFs estimated for each Group in the Events condition. Shaded areas represent the SEM. C. Bar graphs showing the average reconstruction accuracy of the decoding model (Pearson’s r), in each Group. Asterisk (*) indicate a significant difference between ADHD and Control groups (p < 0.05). Error bars represent the SEM.
Eye-gaze results.
A. Pie chart representing the average amount of time that participants spent looking at different areas within the VR classroom. B. Distribution of the proportion of gaze-time towards the teacher (top) and number of gaze-shifts performed away from the teacher (bottom), for all participants in the AD(H)D and Control groups. C. Bar graph represent the average number of gaze-shifts performed away from the teacher, shown separately for the AD(H)D and Control groups, and for the Quiet and Events conditions. No significant differences were found in any comparison (n.s).
Event Related Potentials in response to sound-events.
A. Grand-average ERP to sound-events, shown separately for the AD(H)D vs. Control groups. ERPs are averaged across the cluster of electrodes where significant differences were found (see panel B). The black horizontal line indicates the time-window where significant differences between groups were found (75-155ms). Shaded areas around the waveforms represent the SEM. B. Scalp topographies of the N1 and P2 responses in the AD(H)D and Control Groups, and the difference between them. Electrodes where significant differences between groups were found are marked in white (p<0.05, cluster corrected). C. Grand-average ERP to Artificial and non-verbal human event-sound. ERPs are averaged across the cluster of electrodes where significant differences were found (see panel D). The black horizontal lines indicate the time-windows where significant differences between ERP to the two types of sound-events were found (67-178 and 187-291ms). Shaded areas around waveforms represent the SEM. D. Scalp topographies of the N1 and P2 responses to Artificial and Non-verbal human sounds and the difference between them. Electrode where significant differences were found are marked in white (p<0.05, cluster corrected). E&F. Box plots depicting the average N1 and P2 responses, separately for each group (AD(H)D vs. Control) and Event-type (Artificial vs. Non-verbal human). ** p < 0.001; * p<0.05.
Event-Related changes in Skin Conductance (SC).
A&B. Time course of event-related changes in phasic SC following the sound-events, shown separately for each group (A) and for the two sound-events types (B). Shaded areas around waveforms represent the SEM. Horizontal line represents the time-windows where significant differences were found. C. Average levels of event-related changes in phasic SC (peak between 2-3sec) shown for each group and for Artificial vs. Non-verbal human sound-events. Error bars represent the SEM. **p<0.01.
Event-related gaze shifts.
AD(H)D A. Bar graph showing the number of gaze-shifts performed in 2-sec epochs following event-sounds vs. control epochs, averaged across groups and sound types. No significant differences were found, suggesting that event-sounds were not more likely to elicit overt gaze-shifts. B. Bar graph showing the number of gaze-shifts performed in 2-sec epochs following each type of sound-event, separately for each group. No significant differences were found in any comparison. Error bars represent the SEM in all bar graphs.
Spectral EEG and Skin Conductance Analysis.
A. Grand-Average Power spectral density (PSD) of the Periodic portion of the EEG signal (after applying the FOOOF algorithm), shows two clear peaks corresponding to the alpha (8-12Hz) and beta (15-25Hz) bands. Shaded areas around waveforms represent the SEM. B. Topographical distribution of the average alpha-power and beta-power peaks, with the clusters of electrodes used to detect personal peaks in each frequency band marked by white circles. C. Average alpha- and beta-power in Group (AD(H)D vs. Control) and condition (Quiet and Events). D. Phasic and E. tonic skin conductance levels in the same groups and conditions. No significant between-group difference was found in any comparison (n.s.). Bar graphs represent the mean values, and error bars represent the SEM.
Correlation matrix.
Heatmap of the pairwise spearman’s correlation coefficient between the different neurophysiological measures included in the multivariate analysis. Red-shading indicates negative correlation values and blue-shading indicate positive values. Asterisks* indicate correlation values that pass a non-corrected threshold for statistical significance, however none survived fdr-correction for multiple comparisons.
Result of the dominance analysis of a multivariate logistic regression, describing the contribution of each measure for predicting whether an individual was in the ADHD or control group.
