Grammaticality judgements of children with and without language delay

We examined the grammatical intuitions of children both with and without language delay, assessed via a task presented on computer. We targeted three grammatical structures often reported as compromised in children with language impairments (copula, articles and auxiliaries). 26 children (8 girls) with language delay were recruited (mean age 4;10, range 3;8�6;0). These children met the standard criteria for Specific Language Impairment and underwent an intervention focusing on the three targets. The intervention supplied negative evidence for target omissions, according to the precepts of the Direct Contrast hypothesis (Saxton, 1997). Each child received 20-minute therapeutic sessions daily over a six week period. Children with language delay were tested at four points: pre-intervention, mid-intervention, immediately postintervention and again six months later. To provide a basis for comparison, we also recruited a group of 116 typically developing children (62 girls) (mean age 5;9, range 3;1-7;9). The grammaticality judgement task yielded two measures: (1) judgement of correctness; and (2) reaction time. Although clear differences were found between typical and atypical children, it was clear that even among the oldest typical children, none performed at ceiling. We also found significant improvements in the performance of language delayed children after the intervention