Spelling ability and dyslexia: the case of lexical boundaries knowledge

Spelling ability and dyslexia: the case of lexical boundaries knowledge

First Author: Agnès COSTERG -- Université de Sherbrooke
Additional authors/chairs: 
Daniel Daigle
Keywords: Developmental dyslexia, Orthographic representations, Spelling Ability
Abstract / Summary: 


The ability to read and write correctly includes the construction of orthographic representations well defined (Apel, 2009; Bosse et al., 2014). Recent studies, however, reveal that dyslexic children do not seem to have good representations of lexical boundaries (beginning and end of words) (Daigle et al., 2016; Plisson et al., 2013). The purpose of this study is therefore to evaluate lexical boundaries knowledge in primary school students with dyslexia.


French-speaking dyslexic children (DYS) aged 11.39 years were compared to 18 reading-age matched controls (RA) who were two years younger, to 18 chronological-age matched controls (CA) and to 18 spelling-age matched controls (SA) who were four years younger. These three control groups had no difficulties in reading and writing. Three tasks (lexical decision, lexical identification and lexical permutation) were completed by all participants in oral and written form. The lexical decision included 48 items, half of which corresponding to a single word (cage) and the other half to pseudowords or two words (aveccrème - with cream). The lexical identification consisted in counting the number of words in 18 sentences. For lexical permutation, children had to swap the first and the last word in 12 sentences. Success for each task and each group was calculated.


Results reveal a significant delay in the developmental path linked to the knowledge of lexical boundaries in dyslexic subjects. Indeed, dyslexics are weaker than CAs and RAs, but are comparable to SAs.


It could be interesting, with other visual-orthographic phenomena than lexical boundaries, to reproduce this study with dyslexic students taking into account a triple pairing (RA, CA and SA) in order to verify whether the developmental model, for the phenomena studied, corresponds to a delay (as for lexical boundaries knowledge) or a deviance.