Dynamic assessment of word learning as a predictor of vocabulary, reading comprehension and risk status for the poor comprehender reading profile

Abstract

The Poor Comprehender (PC) reading profile is characterised by difficulty comprehending text despite age-appropriate decoding skills. Risk for this profile is typically identified through static screening instruments measuring pre-existing knowledge, which may produce biased estimates for culturally and linguistically diverse children. In contrast, Dynamic Assessment (DA) measures potential to learn new knowledge and has been shown to reduce bias in screening. To date, however, DA has not been used to identify PC reading profile risk status. Adopting a longitudinal design, we used an adapted DA of word learning to measure growth in vocabulary and reading comprehension among a diverse sample of 322 primary school children (aged 9 years at the first time point) in England over a period of 16-19 months, and to classify later PC reading profile risk status. Two separate factor scores representing phonological and semantic aspects of word learning predicted unique variance in the growth of vocabulary knowledge and reading comprehension after accounting for traditional static predictors. The DA of word learning achieved excellent classification accuracy for identifying poor comprehenders, identifying all of the poor comprehenders with EAL when added to the static assessments. Results suggest that DA of word learning may be a promising tool in screening for the PC reading profile, particularly for EAL pupils, ensuring that the risk of reading comprehension difficulties does not go unidentified in such learners.