Specific language impairment and developmental dyslexia: What are the boundaries? Data from Greek children

• Comparisons between SLI and DD vs. chronological-age and reading-level controls.
• Significance & frequency of phonological vs. non-phonological deficits in SLI & DYS.
• Greater evidence of significant & frequent deficits in reading comprehension in SLI.
• Greater evidence of significant & frequent deficits in decoding skills in DYS.
• Greater evidence of significant & frequent deficits in phonological STM in SLI.

This study examines the significance (between-groups comparisons) and frequency (within-group analyses) of deficits in developmental dyslexia (DD, mainly deficits in decoding and phonemic awareness), specific language impairment (SLI, mainly deficits in listening comprehension), or both (mainly deficits in phonological short-term memory [STM]). Participants included two groups of children who had received a diagnosis of either SLI (N = 15) or DD (N = 15). For the between-groups comparison, the groups were matched pairwise on nonverbal IQ to 30 chronological age controls (CAC) and 30 reading level controls (RLC). For the within-group analyses, the participants were compared to 91 CACs and 63 RLCs. We developed tasks not used for the diagnoses to assess phonological skills (decoding, phonemic awareness, phonological STM) and non-phonological skills (listening and reading comprehension). SLI children scored lower than both DD children and RLCs on tasks assessing listening and reading comprehension, and lower than RLCs on phonological STM and phonemic awareness. Within-group comparisons showed that a higher proportion of SLI than DD children presented severe deficits in the same four domains. The opposite pattern was found for decoding skills (7 SLI children with a severe deficit, versus 13 in the DD group). These findings are discussed in the light of models explaining the overlap between SLI and DD. They highlight the need to assess both phonological and non-phonological skills in SLI and DD children, using both between- and within-groups designs.