“…Several methods were used in neuroimaging studies to determine the coupling direction between neural signals, such as Granger causality analysis (GCA) (Granger, 1969; Im et al, 2010; Jiang et al, 2015), phase transfer entropy (PTE) (Cao, Wang, Liu, & Alexandrakis, 2018; Hillebrand et al, 2016; Lobier, Siebenhühner, Palva, & Palva, 2014; Urquhart, Wang, Liu, Fadel, & Alexandrakis, 2020; Wang et al, 2017), permutation conditional mutual information (Abásolo, Escudero, Hornero, Gómez, & Espino, 2008; Hall & Sarkar, 2011; Li & Ouyang, 2010; Liang, Liang, Wang, Ouyang, & Li, 2015; Wen et al, 2016), and so on. In this study, the GCA method was used, because the GCA is easy to implement, and has been widely used in previous fNIRS studies to estimate the causal relationship between fNIRS time series data, such as cooperation (Cui et al, 2012), teaching (Pan, Novembre, Song, Li, & Hu, 2018), and imitation (Holper, Scholkmann, & Wolf, 2012).…”