The Global and Local Patterns of Reading-Related Cognitive and Ecological Variables in Chinese First-Grade Children

A Cross-Sectional Network Analysis

Bibliographic Data

ID21349986
AuthorsZiqian Wei (0009-0005-1352-887X, The Education University of Hong Kong, China), Lei Wang (0000-0001-5326-565X, University of Oslo), Zhengye Xu (0000-0001-6738-8060, The Education University of Hong Kong, China), Lirong Luo (The Education University of Hong Kong, China), Xinyong Zhang (Guangdong University of Foreign Studies), Ning Li (0000-0003-4098-1838, Guangdong Polytechnic Normal University), Phil Duo Liu (0000-0002-2352-2616, The Education University of Hong Kong, China, corresponding author)
Year2025
Volume58
Issue5
Pages390-406
Publication date2025-09-01
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueJournal of Learning Disabilities (JOURNAL)
Journal identifiersISSN: 0022-2194 • E-ISSN: 1538-4780
PublisherSAGE Publications (PUBLISHER • US)
DOI10.1177/00222194241283198
PMID39376017
OpenAlexW4403225929
LanguageEN
References cited87

In the current study, we tested a network model of reading difficulty by using state-of-the-art psychological network analysis. Four hundred and fifty-three Chinese first-grade children (about 38% female, mean age = 7.00, SD = 0.41) were divided into good ( n = 154), competent ( n = 147), and struggling ( n = 152) readers based on their scores of Chinese character reading. The Extended Bayesian Information Criterion graphical lasso (EBICglasso) method was applied to estimate cross-sectional networks for the three groups. Each network included four cognitive nodes (homophone awareness, morphological structure awareness, phonological awareness, and vocabulary) and two ecological nodes (family socioeconomic status and the number of books at home). Chronological age and nonverbal intelligence were also included in the estimated networks. The global (i.e., global structure and global connectivity) and local patterns (i.e., the most important edges and nodes) in each network were reported. The network comparison results showed that global connectivity was significantly lower among struggling readers than for good readers, implying that a holistic impairment of bidirectional connections among multiple variables relates to the difficulty in learning to read. The theoretical and empirical implications and the significance of applying the network approach to reading research are discussed

Cognition · Cognitive psychology · Correlation · Developmental psychology · Linguistics · Literacy · Partial correlation · Phonological awareness · Reading (process) · Socioeconomic status · Cognitive Science and Mapping · Demography · Functional Brain Connectivity Studies · Mathematics · Mental Health Research Topics · Psychology

Tools

Open DOIOpen Access