Abstract
A Portuguese sample of 50 children with developmental dyslexia (DD) and 50 typical readers (TR) who were matched for age (8–12 years old) were tested on measures of working memory. Relative to the TR, the children with DD performed significantly worse on phonological loop (PL) and central executive (CE) tasks; however, they exhibited no impairments on visuospatial sketchpad (VSSP) tasks. After controlling for the influence of the PL, the group differences in CE tasks were no longer significant. The results of a receiver-operating characteristics curve analysis and a binary logistic regression analysis suggested that the PL and CE tasks (but not the VSSP tasks) were relevant variables for identifying children with DD. Hierarchical linear regression analyses showed that the PL and CE (Backward Digit Span only) tasks were significant predictors of reading and spelling abilities.
Notes
1The reliability of the Digit Span subtest was .80 (split-half), with a test–retest correlation coefficient of .72 (Wechsler, Citation2003).
2The BANC is a comprehensive assessment instrument tapping different functions of children's neuropsychological development, which included 16 tests organized in six main domains: memory (Verbal Learning Test, Narrative Memory, Memory of Faces, Rey Complex Figure Test, and Corsi Block Test), language (Phonological Awareness, Instruction Comprehension, and Rapid Naming), attention and executive functions (Cancellation, Trail, Semantic Verbal Fluency, Phonemic Verbal Fluency, and Tower), motricity, laterality, and orientation. The BANC (Simões et al., Citationin press) was normed on a representative and stratified sample of 1,104 Portuguese children (aged 5 to 15 years) and revealed adequate psychometric properties (e.g., confirmatory factor analysis yielded an adequate model fit with Comparative Fit Index [CFI] = .965 and root mean square error of approximation [RMSEA] = .044 for children aged 7 to 9 years, and CFI = .966 and RMSEA = .046 for children aged 10 to 15 years).