Children with specific language impairment (SLI) lag behind children with typical language development (TL) in their grammatical development, despite equivalent early exposures to a form of adult grammatical corrections in conversation, called recasts (Fey et al., 1999), and their demonstrated ability to learn from recasts in intervention as quickly as do children with TL (Nelson et al., 1996). Recasts are immediate adult responses to child utterances that repeat some of the child's words and correct or otherwise modify the morphologic or syntactic form of the child's prior utterance, while maintaining the central meaning of the child's production. For example, if a child said """"""""It falled down"""""""", an adult corrective recast would be """"""""It fell down"""""""". From one theoretical view, the basis of the grammatical problems of children with SLI involves limitations in working memory, and processing speed and capacity (Leonard et al., 2007). Leonard et al. suggest that intervention approaches that decrease processing demands and offer multiple repetitions of words and sentences have the potential to override processing limitations. The facilitative properties of recasts and the flexibility of their use in treatment give them the capability to fulfill these conditions. The question remains, however, how can clinicians employ recasts in intervention most efficaciously? Warren et al. (2007) suggest that it is time for the field of child language to begin specifically investigating components of treatment intensiveness, such as dosage, and their impact on treatment efficacy. Currently, such research is virtually nonexistent. The proposed study will begin to fill this gap through examination of two specific aspects of dosage: recast rate and recast distribution.
The specific aim of this study is to determine how rate and distribution of recasts affects acquisition of irregular past tense of novel verbs, by 24 children with SLI (4- and 5-yr olds) and 24 younger, language equivalent children with TL (3-yr olds). This study, which utilizes a multifactor between groups design, includes two independent experimental tasks. Experiment 1 examines the children's learning of two irregular past tense novel verbs (dake-doke and kig-kug) in response to one of three rates of recasts, 0.5/min, 1.0/min, or 1.5/min, over five sessions. Experiment 2 examines the children's acquisition of two different irregular past tense novel verbs (jare-jore and twink-twank) when given 20 verb exposures each in 1, 2 or 5 sessions. Outcome variables include percent of accurate production in obligatory contexts of the target verbs during the sessions and on post-test probes. This research proposal represents the fourth study in a path of research that has three long- term goals: (1) establishment of efficacious early intervention techniques that will contribute to the amelioration of the long-term interpersonal and academic consequences of SLI; (2) a detailed understanding of language processing limitations that affect morphosyntactic acquisition of children with SLI; and (3) development of a dynamic assessment protocol that distinguishes late-talking toddlers with SLI from those with TL. The methodology and results of the proposed study will contribute significantly to future randomized clinical trials.
This project will offer specific information about dosage for a widely-used intervention approach, sentence recasting, which facilitates the grammatical development of children with typical language development (cf. Farrar, 1992) and those with specific language impairment (cf. Nelson, Camarata, Welsh, Butovsky & Camarata, 1996). Identification of effective recast dosage parameters might prove to be the key to consistent, effective, intervention outcomes for children with specific language impairment (Warren, Fey & Yoder, 2007). In children with specific language impairment, rapid and early improvement in grammar skills to levels that approximate those of their typically developing peers would be of significant benefit, (Fey & Proctor-Williams, 2000) given the demonstrated long-term interpersonal (cf. Kaiser, Cai, Hancock, & Foster, 2002), and academic difficulties (cf. Catts & Kamhi, 1999) of this population and the associated costs to society. ? ? ?
