Experimental set up and schematic illustration of the procedure followed for analysis 2 to 5.

Top figure shows the experimental set up for solo play. Below, on the left, is shown the procedure followed to parse the looking behaviour and create the variable “attention”, and further cross-correlation analyses (analysis 2 and 5). On the right, instead, is shown the steps followed to identify attentional episodes in the EEG signal and further EEG analyses (analysis 3 and 4).

Descriptive analyses on infant attentional behaviour during the solo play interaction.

(A) Average number of attentive and inattentive looks per minute at 5 months (left) and 10 months (right), (B) Average number of attentive and inattentive looks minute by minute. Asterisks show the significance values of comparisons examining how average number of looks per minute of the interaction differed as a function of age. (C) Average duration spent in one of the two possible attentional states: attentive and inattentive, and (D) minute by minute. Again, asterisks show the significance values of comparisons examining differences as a function of age. (E) Percentage of time infants spent in attentive vs. inattentive states, during the whole interaction. (F) Histogram showing the distribution of the proportion of all the looks that lasted less than or equal to 5 seconds (right) and more than 5 seconds (left) at 5 months (yellow) and 10 months (blue). Continuous black line indicates the mode of each distribution. Significance is indicated with asterisks where * = p<0.05, ** = p<0.01, and *** = p<0.001. Error bars represent SEMs.

ACF and survival probability analyses of the looking behaviour.

(A) Autocorrelation of the time series looking behaviour at 5 months (in yellow) and 10 months (in blue). (B) Survival analysis. Survival probability function for looking behaviour toward object toys. The survival function is the probability that a look will survive a given time. Yellow line shows data from 5-months-old infants with confidence bounds (dotted yellow line) and blue line shows data form 10-months-old infants with confidence bounds (dotted blue line). (C) Derivative of the Survival Probability at 5 months (yellow) and 10 months (blue).

Relationship between infant autonomic arousal and attention.

Autocorrelation results for infant autonomic arousal at 5 months (A) and 10 months (C). Significant clusters are indicated by red dots. Cross correlation between infant autonomic arousal and attention at 5 months (B) and at 10 months (D). Infant autonomic arousal forward-predicting infant attention on the negative lags, infant attention forwardpredicting infant autonomic arousal on the positive lags. Black lines show the cross-correlation values, shaded grey areas indicate the SEM. Shaded yellow areas show confidence intervals from the permuted data. Significant time lags identified by the cluster-based permutation analyses are shown by a thick red line.

Calculation of theta power changes around an attention episode.

Results of the linear mixed effects models conducted to examine whether individual looks accompanied by higher theta power are longer lasting. For each look, we calculated the association between the total duration of the look and relative theta power during five time-windows (-2000msec to -1000msec and -1000msec to 0 prior to the look, and 0 to 1000msec, 1000 to 2000msec and 2000 to 3000msec before the look), using a series of separate linear mixed effects models. (A) Shows results at 5 months where the y-axis is the t value, and (B) shows the results at 10 months. Asterisks (*) indicate p values < .05. Central channels include: ‘FC1’, ‘FC2’, ‘C3’, ‘Cz’, ‘C4’, ‘CP1’ and ‘CP2’.

Calculation of theta power changes within an attention episode.

Bar plots for the average relative theta power throughout a look, at both time points (5 and 10 months) and at different brain networks (central, occipital, and frontal). Asterisks (*) indicate p<0.05. Error bars represent SEMs.

Relationship between infant relative theta activity, infant attention, and infant autonomic activity.

Autocorrelation for infant theta activity at 5 months (A) and at 10 months (D). Significant clusters are indicated by red dots. Cross-correlation results between infant theta activity and infant attention at 5 months (B) and at 10 months (E). Infant theta activity forward-predicting infant attention on the negative lags, infant attention forward-predicting infant theta activity on the positive lags. Cross-correlation results between infant autonomic activity (indexed by heart rate activity) and relative theta power at 5 months (C) and at 10 months (F). Infant autonomic activity forward-predicting infant theta activity on the negative lags, infant theta activity forward-predicting infant autonomic activity on the positive lags. Black lines show the Spearman correlation at each time lag, shaded grey areas indicate the SEM. Shaded yellow areas show confidence intervals from the permuted data. Significant time lags identified by the cluster-based permutation analyses are shown by a thick red line.