Impact of Phonological Awareness on Vocabulary Acquisition in Dyslexic Children: Towards an Artificial Intelligence Model for Early Diagnosis

This article examines the influence of phonological awareness on vocabulary acquisition among dyslexic children in a multilingual Moroccan context. The study compares the performance of 20 children, both dyslexic and non-dyslexic, aged 6 to 16 years, assessed using tests for phoneme segmentation, deletion, and blending to measure phonological awareness, as well as receptive and expressive vocabulary tests. The results reveal that dyslexic children score significantly lower in phonological awareness, confirming the presence of phonological deficits within this group. However, correlation and regression analyses indicate weak and non-significant relationships between phonological awareness and vocabulary, contrary to existing literature. This divergence may be attributed to the small sample size, age dispersion, the use of non-adaptive standardized tests, and the complexity of the Moroccan linguistic context. To address these limitations, the development of a supervised classification model using a neural network is proposed. This model, based on local data, could refine diagnostics and identify diverse phonological and lexical profiles, thereby facilitating the development of tailored educational interventions. This study highlights the need for contextualized assessment tools and innovative approaches, integrating artificial intelligence, to better understand and support vocabulary acquisition in dyslexic children within multicultural environments.