Neurocognitive constructs as longitudinal predictors of reading fluency

Reading fluency (RF) involves the automaticity of many distinct reading skills (e.g., pacing, word recognition, phonological awareness) and allows cognitive resources to be allocated to higher-order reading skills (e.g., comprehension, synthesis). Early identification of students at-risk for RF deficits is critical, but many screeners require a level of basic proficiency with foundational reading skills. This review examines the longitudinal relationships between neurocognitive precursors of RF with special consideration of cognitive processing demands of measurement tasks. Eighteen articles yielded 80 unique correlations between 6 neurocognitive constructs (phonological processing, rapid automatic naming, working memory, nonverbal reasoning, orthographic processing, and visual-spatial processing) measured through 35 unique subtests/tasks and 2 primary RF outcomes (word RF and text RF) measured through 11 unique subtests/tasks. Phonological processing was the most frequently assessed neurocognitive predictor (41.3% of total correlations), followed by rapid automatic naming (32.5%). Nearly all correlation coefficients were significant (83.75%); 28.75% of correlations suggested a strong predictive relationship between a neurocognitive predictor and an RF outcome. The strength of correlation largely varied as a function of the task demands as evidenced through the supplemental process analyses, suggesting future consideration should be given to the specificity of task selection to improve ecologically valid conclusions. Evidence from this review supports the need to consider specific neurocognitive functions, specifically rapid automatic naming and orthographic processing, as critical components of dynamic and comprehensive assessments of reading to help identify students at-risk for future RF challenges.