Predictors of reading skills in Turkish-speaking fourth grade children

Abstract

The present study aimed to investigate the concurrent contributions of phonological awareness (PA), rapid automatized naming (RAN), vocabulary, morphological awareness (MA), processing speed (PS), working memory (WM), and inhibitory control (IC) to reading skills in Turkish. It also examined the complex network of relationships among the variables by focusing on RAN, text reading fluency (TRF) and reading comprehension (RC). In the data collection process, a variety of cognitive and linguistic measures were administered to 112 Turkish-speaking Grade 4 children. Based on previous research, a preliminary model of reading was developed and tested through multiple regression analyses as part of a classical path analysis. The results showed that IC was the strongest predictor of RAN, followed by PS. In the next layer of the analysis, RAN explained the largest amount of variance in TRF, followed by MA and PS. In the ultimate model, MA was the most powerful predictor of RC, followed by PA, TRF and vocabulary. MA also explained a considerable amount of variance in vocabulary, which highlighted the interface between morphological and semantic processes involved in RC. Additional analyses revealed that the predictor variables also made a multitude of indirect contributions to RAN, TRF and RC in the highly agglutinative and morphologically rich structure of Turkish.