Semantic and inferencing abilities in children with communication disorders.

BACKGROUND Semantic and inferencing abilities have not been fully examined in children with communication difficulties. AIMS To investigate the inferential and semantic abilities of children with communication difficulties using newly designed tasks. METHODS & PROCEDURES Children with different types of communication disorder were compared with each other and with three groups of typically developing children: those of the same chronological age and two groups of younger children. In total, 25 children aged 11 years with specific language impairment and 22 children, also 11 years of age, with primary pragmatic difficulties were recruited. Typically developing groups aged 11 (n=35; age-match), and those aged 9 (n=40) and 7 (n=37; language similar) also participated as comparisons. OUTCOMES & RESULTS For Semantic Choices, children with specific language impairment performed significantly more poorly than 9- and 11-year-olds, whilst the pragmatic difficulties group scored significantly lower than all the typically developing groups. Borderline differences between specific language impairment and pragmatic difficulties groups were found. For inferencing, children with communication impairments performed significantly below the 11-year-old peers, but not poorer than 9- and 7-year-olds, suggesting that this skill is in line with language ability. Six children in the pragmatic difficulties group who met diagnosis for autism performed more poorly than the other two clinical groups on both tasks, but not statistically significantly so. CONCLUSIONS Both tasks were more difficult for those with communication impairments compared with peers. Semantic but not inferencing abilities showed a non-significant trend for differences between the two clinical groups and children with pragmatic difficulties performed more poorly than all typically developing groups. The tasks may relate to each other in varying ways according to type of communication difficulty.