Eye contact and emotional face processing in 6-month-old infants: Advanced statistical methods applied to event-related potentials

Event-related potential (ERP) studies with infants are often limited by a small number of measurements. We introduce a weighted general linear mixed model analysis with a time-varying covariate, which allows for the efficient analysis of all available event-related potential data of infants. This method allows controlling the signal to noise ratio effect on averaged ERP estimates due to small and varying numbers of trials. The method enables analyzing ERP data sets of infants, which would often not be possible otherwise. We illustrate this method by analyzing an experimental study and discuss the advantages in comparison to currently used methods as well as its potential limitations. In this study, 6-month-old infants saw a face showing a neutral or an angry expression in combination with direct or averted eye gaze. We examined how the infant brain processes facial expressions and whether the direction of eye gaze has an influence on it. We focused on the infant Negative Central ERP component (Nc). The neutral expression elicited larger amplitude and peaked earlier than the angry expression. An interaction between emotion and gaze was found for Nc latency, suggesting that emotions are processed in combination with eye gaze in infancy. © 2009 Elsevier B.V. All rights reserved.