Reading in Arabic: How Well Does the Standard Model Apply?

Abstract

Purpose We investigated the cognitive and linguistic processes that underlie reading in Arabic in relation to a well-defined theoretical framework of reading and the factors that underlie reading. Method The sample was 201 (101 boys, 100 girls) 3rd-grade Arabic-speaking children. Children were administered measures of Vocabulary, Phonological Awareness (PA), Naming Speed, Orthographic Processing, Morphological Awareness (MA), Memory, Nonverbal Ability, and 5 reading outcomes. Hierarchical regression analyses were conducted for each of the 5 reading outcomes to investigate the predictors of children's reading. Results Each of the constructs explained unique variance when added to the model. In the final models, PA was the strongest predictor of all outcomes, followed by MA. In a follow-up analysis, participants were divided into good and poor decoders, based on their Pseudoword Reading scores. Good decoders outscored poor decoders on every measure. Within-group regression analyses indicated that poor decoders relied on more component processes than good decoders, suggesting a lack of automaticity. Variance in reading outcomes was better predicted for poor decoders than for good decoders. Conclusion These results indicate that standard predictors apply well to Arabic, showing the particular importance of PA and MA. Longitudinal and instructional studies are required to determine developmental patterns and ways to improve reading performance.