The Foundations of Literacy Development in Children at Familial Risk of Dyslexia

Abstract

The development of reading skills is underpinned by oral language abilities: Phonological skills appear to have a causal influence on the development of early word-level literacy skills, and reading-comprehension ability depends, in addition to word-level literacy skills, on broader (semantic and syntactic) language skills. Here, we report a longitudinal study of children at familial risk of dyslexia, children with preschool language difficulties, and typically developing control children. Preschool measures of oral language predicted phoneme awareness and grapheme-phoneme knowledge just before school entry, which in turn predicted word-level literacy skills shortly after school entry. Reading comprehension at 8½ years was predicted by word-level literacy skills at 5½ years and by language skills at 3½ years. These patterns of predictive relationships were similar in both typically developing children and those at risk of literacy difficulties. Our findings underline the importance of oral language skills for the development of both word-level literacy and reading comprehension.

Keywords: dyslexia; familial risk; language impairment; language skills; phonological skills; reading comprehension; reading development.

Fig. 1.

Two-group structural equation path model showing the longitudinal relations between Time 1 (preschool) measures of language and speech, Time 2 preliteracy skills, Time 3 word-level literacy skills, and Time 5 reading comprehension for the at-risk and control samples. Ellipses indicate latent variables, and rectangles indicate observed variables. On each path, values for the at-risk group are outside parentheses, and values for the typically developing (control) group are inside parentheses. Values on single-headed arrows from the latent to the observed variables are standardized factor loadings (or standardized regression weights for reading comprehension), and values on single-headed arrows between the latent variables are standardized regression weights. Double-headed arrows indicate correlations (covariances). Solid lines indicate statistically significant relationships, and dashed lines indicate statistically nonsignificant relationships.