Learning to Read in Chinese: A Meta-analytic Structural Equation Modeling Analysis of Cognitive-Linguistic Skills in Chinese Word Reading

Abstract

This paper reports the first meta-analytic structural equation modeling on the associations of lower-level cognitive-linguistic skills (rapid automatized naming/RAN, orthographic awareness/OA, phonological awareness/PA, morphological awareness/MA, and vocabulary) with word reading among native Chinese-speaking children, based on pooled correlational matrices (k = 446; N = 70,734). Moderator analyses explored how the (relative) associations of these skills varied across readers of different Chinese scripts, literacy-instruction contexts, grade levels, and word reading tasks. We found that (1) all skills were significantly and independently associated with word reading with the meaning-based pathways (MA and vocabulary) playing a more salient role than code-based pathways (OA and PA); (2) children reading the traditional script showed stronger associations of OA and MA with word reading than those reading the simplified script; (3) contexts involving phonological instruction were associated with a stronger role of PA but a reduced role of MA in word reading; (4) developmentally, there was a decreasing association of code-based strategies but an increasing association of meaning-based strategies (i.e., increased effects of MA and vocabulary) with word reading; and (5) the association of RAN was stronger with word reading fluency than with accuracy, whereas a converse pattern was found for that of PA. These findings pointed to interrelated patterns of cognitive-linguistic skills in Chinese word reading. They suggested universal and language/script-specific processes in reading acquisition, while also underscoring the differential functioning of component processes in reading acquisition across contexts, developmental stages, and reading tasks.