*2.7. Baseline Normalisation*

AWI, MI, HR and EI signals were epoched according to the narrative sequences and z-score transformed with respect to the BSL as [76]:

$$\mathbf{x}'(t) = \left[\mathbf{x}(t) - m\_{\rm BSL}\right] / s\_{\rm BSL} \tag{1}$$

where *x*(*t*) is the z-score transformed signal, *x*(*t*) is the original signal, *mBSL* is the temporal mean of *x*(*t*) in the BSL epoch and *sBSL* is the temporal standard deviation of *x*(*t*) in the BSL epoch.

Then, the signals were temporally averaged across each narrative sequence in order to obtain a condensed stimulus-related index [80]. For each Video × Theme combination, outliers were identified by means of the inter-quantile range (IQR) criterion as points outside the interval [Q1 − 1.5 × IQR; Q3 + 1.5 × IQR], where Q1 is the first quartile, Q3 is the third quartile and IQR = Q3 − Q1 [81].
