Abstract
Recent Event Related Potentials (ERPs) studies using eye-tracking and a gaze contingent approach to control gaze fixation location have shown that the face-sensitive N170 ERP component was modulated when participants’ gaze was fixated on the different facial features of upright or inverted faces (Nemrodov, et al. 2014; Itier & Preston, 2018). In the first experiment we re-analyzed these data sets with Mass-Univariate statistics using the LIMO toolbox. These statistics are more resistant to Type I & II errors and allow the investigation of all time points and electrodes (Pernet, et al. 2011). The results revealed a significant effect of fixation location, orientation, as well as their interaction. These significant differences were widespread across the scalp and were maximal before and after the N170 ERP component. Both main effects were largest during the P1 to N170 and N170 to P2 windows. The face inversion effect was maximal for the nasion fixation around 130ms, i.e. about 20ms before the N170 peak. In the second experiment (unpublished) the exact same experimental paradigm was used, except the images were of upright and inverted houses. The results showed a fixation and orientation effect of houses as well as a small interaction between the two. However, these effects occurred over a much smaller time window and were considerably smaller in magnitude with F values much lower than those seen with faces. These results suggest that face and house processing are differentially impacted by orientation and fixation location. Critically, most of these processes occurred outside of the N170 peak, emphasizing the need to move towards robust statistics to unravel the time course of face and object neural processing.