Discriminant Analysis of Cognitive Linguistic Indicators for Early Identification of Dyslexia in Nigerian Primary School Pupils

There is a reasonable number of pupils suffering from dyslexia in Nigeria; however, very little has been done to identify and provide targeted support for them. The absence of reliable, context-specific diagnostic tools has made early detection challenging, leaving many children to struggle with persistent reading and learning difficulties. This study applied the Primary Reading Inventory (PRI) model to predict and identify dyslexic pupils using three cognitive-linguistic indicators: Phonemic Awareness (PA), Rapid Naming and Vocabulary (RNV), and Word Recognition and Meaning (WRM). Discriminant analysis was employed to determine the capacity of these indicators to differentiate between dyslexic and non-dyslexic pupils. The results revealed a strong canonical correlation coefficient (0.88) and a low Wilks’ Lambda value (0.224), indicating a high level of discrimination between the groups. The analysis further showed that Phonemic Awareness was the most influential predictor, followed by RNV and WRM, with the overall model achieving a classification accuracy of 97.2%. These findings demonstrate the reliability of the PRI model as an effective and practical approach for identifying dyslexia within school contexts. The study underscores the importance of integrating statistical approaches with established cognitive-linguistic models to identify dyslexic children across schools in Nigeria and other African cities, thereby promoting inclusive education, facilitating targeted interventions, and fostering equitable learning outcomes for pupils with reading difficulties.