Background: Specific language impairment (SLI) is currently partly defined by the presence of non-verbal IQ scores in the normal range. However, not only is there a debate concerning where ‘normal thresholds’ should be, but increasing information about the presence of processing deficits in SLI have led some researchers to question the use of IQ criteria in clinical diagnosis. In particular, little is known about the longitudinal and developmental patterns of cognitive performance in this population.
Method: Data from a long-term follow-up study of SLI was examined in 82 children defined at original participation as having SLI who had IQ measurements at 7, 8, 11 and 14 years.
Results: Analyses revealed a significant fall between 7 and 14 years of over 20 IQ points. This fall took place mainly between 8 and 11 years but was still continuing between 11 and 14 years. Further investigation revealed different groups of children showing different developmental patterns in IQ, even after controlling for baseline measurement. These groups also showed significantly different language outcomes at 14 years. Analyses controlling for IQ at 7 were also performed that suggested a dynamic process between language and cognitive development.
Conclusions: These findings appear to be in agreement with a model of impairment that views ‘Residual Normality’ as unlikely (Karmiloff-Smith, 1998; Thomas & Karmiloff-Smith, 2002). The implications are therefore discussed in relation to the dynamic development of systems along with the possible cognitive mechanisms (such as working memory) that might interact with language to create an SLI profile.