Automatic extraction of meaning from visual number symbols detected by frequency-tagged EEG in children

Abstract

AbstractSymbolic number representation and manipulation is key for successful mathematical learning. However, the mechanism by which at some point in development number symbols (i.e.,1,2,3, etc.) begin to automatically elicit useful meaning remains unresolved. Previous evidence highlighted that it is not possible to ignore the numerical magnitude when looking at number symbols, at least for adults. However, the neural mechanism behind the progressive automatization of symbol processing remains largely unknown, namely because these kinds of cognitive processes are difficult to isolate due to the general cognitive skills involved in any explicit task design. We thus developed an experimental paradigm specifically targeting the neural correlates of implicit magnitude representations by frequency-tagging magnitude changes within a visual stream of digits. Automatic magnitude processing was assessed by presenting a stream of number symbols with a frequency-tagged change of magnitude allowing to identify automatic categorization of the symbols by their magnitude in (pre)school-aged children. Stimuli were displayed with a sinusoidal contrast modulation at the frequency of 10 Hz and Steady-State Visual Evoked Potentials were recorded. These electrophysiological measurements showed a neural synchronization at the harmonics of the frequency of the magnitude changes recorded on electrodes encompassing bilateral occipitoparietal regions. The current findings indicate that magnitude is a salient semantic feature of the number symbols, which is deeply associated to digits in long-term memory across development.Significance StatementAcquiring strong semantic representations of numbers is crucial for future math achievement. However, the learning stage where magnitude information becomes automatically elicited by number symbols (i.e., Arabic digits from 1 to 9) remains unknown, namely due to the difficulty to measure unintentional automatic processing of magnitudes. We used a new experimental paradigm especially targeting the neural mechanisms involved in the automatic processing of magnitude information conveyed by number symbols. Frequency-tagged electrophysiological responses have the advantage to provide large amounts of reliable data with a high signal-to-noise ratio in a minimal amount of time. The current study is the first to take advantage of this in developmental populations to understand early automatic magnitude representations in children’s numeracy development. The electrophysiological responses demonstrate that the magnitude information is already automatically accessed from number symbols in children at the end of preschool, highlighting the importance of the first years of life for building automatized magnitude processing skills.