Early Predictors of Hebrew Reading Development: A Multicomponent Model from Kindergarten Through First and Fourth Grades

Purpose: While most longitudinal research on pre-reading predictors has focused on the transition from kindergarten to Grades 1–2, less is known about their long-term predictive power as children progress from phonological decoding to proficient word recognition. This 5-year study investigated kindergarten language-related predictors of word-reading accuracy and fluency trajectories from initial to advanced stages, examining developmental changes in their relative importance within Hebrew’s unique characteristics, including its dual writing versions–pointed-transparent and unpointed-deep–and its rich Semitic morphology. Method: A total of 515 Hebrew-speaking children (55.3% girls) from northern Israel was assessed on phonological awareness (PA), rapid automatized naming (RAN), letter knowledge (LK), and morphological awareness (MA) in kindergarten (mean age=5.9 years). Word-reading accuracy and fluency were measured in Grade 1 using pointed script, and in Grade 4 using both pointed and unpointed versions. Results: Structural equation modeling revealed that early PA predicted word-reading abilities only in initial stages, while early MA specifically contributed to advanced stages, particularly in unpointed script. Early LK and RAN predicted word-reading abilities in both grades, with sustained direct effects on Grade 4 word-reading fluency across both versions and accuracy in unpointed script. Conclusion: These findings demonstrate that kindergarten language-related skills, despite their early origin, not only establish fundamental reading abilities but also directly support advanced reading proficiency, underscoring the importance of comprehensive kindergarten programs that foster both early and long-term reading success. The distinct predictive patterns observed across reading stages, demands, and orthographic contexts enhance theoretical understanding of general reading development and script-specific frameworks.