• Citer

Neural entrainment during visual statistical learning in infancy

par Capparini, Chiara ;Dontaine, Pauline ;Wens, Vincent ;Bourguignon, Mathieu ;De Tiege, Xavier ;Aeby, Alec ;Bertels, Julie
Référence International Convention of Psychological Science (ICPS) (Bruxelles, Belgium)
Publication Publié, 2023-03

Poster de conférence

• Détails
• Contenu

Résumé :

Our visual world includes a large amount of complex and noisy information. Visual statistical learning refers to the ability to detect and extract regularities from the environment. This learning mechanism has been observed even in newborns and allows to organise visual stimulation in a coherent representation (Bulf et al., 2011). Thus far, statistical learning has been measured with post-exposure behavioural tasks that only measure the outcome of learning. Further, behavioural tasks administered to infants may lead to ambiguous interpretations since there is no clear consensus about the directionality of the expected learning outcome (i.e., novelty or familiarity effect). Electrophysiological measures can be acquired while learning occurs and can shed light onto the temporal course of learning. At present, investigations of the ongoing learning processes during the exposure phase have been limited to the auditory domain (Choi et al., 2020). In this exploratory study, we propose an EEG frequency-tagging approach to study infants’ neural entrainment in response to visual regularities. Participants are 4- to 6-month-old full term infants. The expected sample size is 10-15 participants per condition (depending on the amount of usable data). Infants are presented with a continuous stream of 8 colourful shapes appearing in the centre of the screen at a frequency of 6 Hz. Shapes are organised in doublets repeatedly presented in a pseudo-random order. To make sure that the infant brain has encoded the doublet presentation, the standard doublet condition is compared with a control doublet condition, in which only the first element of the pair follows the rule, and to a control single condition, in which individual shapes are randomly presented. Infants are randomly assigned to one of the three conditions before the beginning of the experiment in a between-subject design. Each trial lasts 20 seconds and the inter-trial interval includes the presentation of an attention-getting video of variable duration. Trials are presented until the participant is attentive and in an alert state. The analysis plan includes a standardised pre-processing step of the EEG data (Bigdely-Shamlo et al., 2015). Following this, neural entrainment at the frequency of visual stimulation (6 Hz) and at the doublet frequency (3 Hz) is compared across conditions. If the regularities in the stream of shapes are detected, we hypothesised not only a neural activation at 6 Hz but also a progressive response at 3 Hz for the infants assigned to the standard doublet condition. At the same time, responses at 6hz are expected in each condition and will confirm that the infant has been looking at the visual stream of stimuli. These results will be crucial to better understand early learning skills as they unfold during stimulus exposure. Also, they may provide a neural signature of the emerging statistical learning abilities that past research has only reported using post-exposure behavioural tasks.