Responsiveness of children with specific language impairment and maternal repairs during shared book reading

Background: Previous studies indicate that (1) children with specific language impairment (SLI) produce a high number of problematic answers during shared book reading, and that (2) adult communicative input has a key role in supporting child responsiveness. However, little is known about the strategies used by parents to repair children's inadequate answers and their effectiveness in supporting appropriate responses. Aims: Repairs (any utterance aimed to correct a child's problematic answer) produced by mothers of children with SLI and mothers of typically developing children during shared reading conversation were compared: (1) to examine whether there are significant differences in their features, and to what extent these differences (if any) are accounted for by limitations in child responsiveness and language abilities; and (2) to assess whether-and in which cases-maternal repairs are effective in eliciting appropriate answers from children with SLI. Methods & Procedures: Ten preschool-aged children with SLI, ten age-matched typically developing children, and ten mean length of utterance-matched typically developing children with their mothers participated in the study. Each mother-child dyad was videotaped during four sessions of shared book reading at home. Each maternal repair was coded according to the level of support (i.e., presence of crucial information and familiar topic) provided to elicit the child's correct answer. Each child's answer following a maternal repair was coded on the basis of both linguistic production and content appropriateness. Outcomes & Results: Mothers of children with SLI produced significantly more high-supportive repairs than mothers of age-matched children, but not more than mothers of mean length of utterance-matched younger children. Sequential analysis applied to maternal repairs and children's answers showed that supportive repairs significantly affected the occurrence of minimally acceptable answers produced by children with SLI, while non-supportive repairs affected significantly the occurrence of inadequate answers. Children with SLI behaved in a similar way to mean length of utterance-matched younger children, showing spontaneous motivation to imitate crucial information included in high-supportive repairs. Conclusions & Implications: The findings suggest that mothers of children with SLI adjust their repairs to their children's linguistic limitations. The use of supportive repairs is functional to model the children's answers, enables them to practise new words, and contributes to the children's experience of being an active interlocutor. These findings have a potential clinical value that can be used in language intervention programmes based on shared book reading.