Beyond phonology: Independent and moderating effects of visual-spatial skills on dyslexia classification in a transparent orthography.

Developmental dyslexia (DD) is primarily attributed to phonological processing deficits, but visual-spatial processing may also shape individual variability. Relatively few studies have examined whether discrete CHC-based visual-spatial indicators provide incremental and conditional information for dyslexia classification beyond phonological and RAN-based predictors, particularly in transparent orthographies. This study tested whether visual-spatial indicators have independent and moderating effects on dyslexia classification in a transparent orthography, within a moderation (risk × protection) framework. Participants were 658 Polish-speaking, age- and sex-matched, school-aged students (329 with DD, 329 without DD). Hierarchical logistic regression models contrasted the contributions of phonological composites, RAN, and Cattell-Horn-Carroll visual-spatial (Gv) indicators-perceptual organization (PO), visual memory (VM), and visual-motor coordination (VMC)-to dyslexia status. Phonological and literacy measures were the strongest predictors. However, visual-spatial indicators showed selective effects, with perceptual organization emerging as the most robust Gv predictor, whereas VMC showed a significant but cautionary adjusted association consistent with shared-variance/suppression effects. Interaction analyses showed that higher PO and VM moderated the associations between phonological/RAN-based predictors and dyslexia classification, such that these predictors showed reduced classification strength at higher levels of PO or VM. Findings provide evidence that visual-spatial indicators are independently and conditionally associated with dyslexia classification, supporting an interaction-aware, multifactorial account of DD.