Psycholinguistic markers in the spellings of bilingual children with developmental dyslexia

Ramos, S. 1 , Rickard Liow, S. J. 1 , Yap, M. 1 , Ho, S. L. 1, 2 & Tan, J. 1, 2

1 National University of Singapore
2 Dyslexia Association of Singapore

The lack of standardized measures for oral and written language performance in bilingual populations makes the diagnosis of developmental dyslexia particularly difficult (Geva, 2000). In Experiment 1, we compared the performance of 20 bilingual children (aged 9-11 years) identified as having dyslexia against 20 age-matched typically developing bilingual children, from the same English-medium school system in Singapore, on a series of common cognitive linguistic tests: BLAB receptive and CELF expressive vocabulary, Naglieri Nonverbal Ability Test, WRAT-4 Spelling, WMBC verbal and visuo-spatial working memory. Significant differences in the raw scores indicated that several of the children identified as having reading and spelling problems were co-morbid for oral language deficits, such as Specific Language Impairment (SLI) and/or were relatively weak English Second Language learners (ESLs). In Experiment 2, we looked for the significant psycholinguistic markers of language and literacy problems in the same bilingual children using their responses to an experimental spelling task. Multiple regression analyses on absolute spelling accuracy, and on a qualitative measure of spelling accuracy, were conducted using orthographic similarity, word frequency, imageability and length as predictors. These analyses revealed differences in the predictors for spelling: the spelling performance of typically developing children was influenced by word frequency, whereas that for children previously identified as having phonological dyslexia or mixed dyslexia (i.e., phonological processing problems consistent with SLI/ESL deficits in their oral language), was influenced by orthographic similarity, word frequency and imageability, but not by word length, and that of children previously identified as having surface dyslexia was only predicted by word length. The implications of this methodology for in-depth cognitive neuropsychological assessment of bilingual children, and more targeted intervention programmes, are discussed.