Sentence comprehension in children with specific language impairment: effects of input rate and phonological working...

Many children with specific language impairment (SLI) exhibit sentence comprehension difficulties. In some instances, these difficulties appear to be related to poor linguistic knowledge and, in other instances, to inferior general processing abilities. Two processing deficiencies evidenced by these children include reduced linguistic processing speed and diminished phonological working memory (PWM) capacity. While providing children with SLI a slower input rate appears to facilitate their immediate (real-time) processing of language; it is not known whether such a manipulation might also promote better offline (conventional) sentence comprehension. Complicating the potential benefits of a slower input rate on these children's offline comprehension is that a slower input rate might negatively interact with their PWM deficit. That is, a slower presentation rate may have the unintended consequence of placing too great a temporal burden on these children's vulnerable PWM system, thereby leading to even poorer comprehension. That is, a slower input rate presumably requires the listener to store previously processed information for a longer period while new, incoming material is being processed.

To investigate the interaction of input rate and PWM capacity on the sentence comprehension of children with SLI. It was hypothesized that if a slower input rate negatively affects the PWM functioning of children with SLI, these children would comprehend significantly fewer sentences presented at a slower rate versus sentences presented at a normal rate. It was also hypothesized that input rate should have no differential effect on the typically developing control children.

A group of 12 school-age children with SLI, a group of 12 age-matched children (CA), and a group of 12 receptive syntax-matched (RS) children (all with normal-range non-verbal IQ and hearing) completed a conventional picture-pointing sentence comprehension task. Sentences were presented at a normal speaking rate and a slower rate (25% time expanded). Children also completed a non-word repetition task, which served as an index of PWM capacity.

Results of the PWM task revealed that the children with SLI repeated fewer three- and four-syllable non-words than both the CA and RS children, suggesting they had reduced PWM capacity. On the sentence comprehension task, a significant group by input rate interaction emerged. The children with SLI yielded significantly poorer comprehension of normal-rate sentences compared with both the CA and RS children. Whereas input rate had no differential effect on CA and RS children's comprehension, the comprehension of the children with SLI improved to the level of RS children in the slow-rate condition. Correlational analyses showed no significant association between PWM and sentence comprehension in either rate condition for any of the groups.

Together, these findings suggest that the simple manipulation of altering speaking rate to children with SLI indeed can facilitate sentence comprehension, with no apparent detrimental effect on the operations of their PWM system.