Poster de conférence
Résumé : Thus far, most evidence of statistical learning in infancy has relied on post-exposure behavioural paradigms. Nevertheless, the directionality of the behavioural learning outcome (i.e., novelty or familiarity effect) may be hard to predict and, in turn, this may raise interpretative challenges. In this context, electrophysiological measures obtained with a frequency-tagging paradigm can be acquired while learning occurs and can help to better characterize or disambiguate unclear behavioural findings. In the present work, we aimed to compare neural and behavioural evidence of visual statistical learning longitudinally. We investigated early learning skills over the first year of infancy at three timepoints, namely at 3 (T1), 6 (T2), and 9 months of age (T3). Data collection is still ongoing, with T1 concluded and T2 and T3 in progress. The final sample at T1 was a group of 30 (17 females) 3-month-old infants (M = 108 days, SD = 9.6 days). At each timepoint, infants took part in an EEG frequency-tagging paradigm in which they were exposed to a sequence of shapes organised in doublets. Shapes were individually presented at the centre of the screen at a frequency of 6 Hz, hence a doublet appeared at a frequency of 3 Hz. Following a familiarization phase of at least 8 sequences and 120 s of cumulative looking, participants began a test phase in which 4 novel and 4 familiar doublets were presented and looking time was recorded. Preliminary EEG results at T1 revealed that the 3-month-olds’ brain could already detect the visual regularities in the stream of shapes with significant occipital activity at the doublet frequency, especially at the first harmonic of the doublet frequency (9 hz). At the individual level, 60% of the 3-month-old participants showed significant occipital activity at the doublet frequency and its harmonics. On the other side, looking time measures at test phase did not reveal any significant difference between novel and familiar doublets. These results suggested an early-emerging brain mechanism to detect visual statistical regularities during stimulus exposure, which seems already in place at 3 months of age and may be a more robust index than post-exposure behavioural measures. Data collection at T2 and T3 is still ongoing and results will be presented at the conference.