![[ Visit Client Website ]](images/banner.JPG)
Friday, May 8, 2009
Northwest Hall (Chicago Hilton)
11:00 AM
Background: A variety of authors have reported that children with autism show specific deficits in imitation skills (Rogers, Hepburn, Stackhouse and Wehner, 2003). Moreover, individual differences in imitation were shown to predict subsequent gains in language, even if even if initial child characteristics (language age, IQ) were statistically controlled (Stone & Yoder, 2001, Charman, 2003, Toth, Munson, Meltzoff and Dawson 2006).
Objectives: In this longitudinal study, we administered a comprehensive imitation battery that evaluates four types of imitative behaviors: manual, object, oral-motor and verbal (Rogers, Hepburn, Stackhouse and Wehner, 2003). We aimed to evaluate predictive relations between these four types of imitative behaviors and children’s subsequent gains in language skills.
Methods: The sample consisted of 34 individuals with autism (chronological age: mean = 55.9 months; SD = 11.9 months), who participated in the control condition of a randomized controlled intervention trial. At baseline, children were administered the imitation battery. A total of 24 imitation probes were administered over the course of 2 lab visits, with 5-7 tasks in each category. In addition, we evaluated children’s receptive and expressive language abilities using the Mullen Scale of Early Learning (Mullen, 1995). Children’s language abilities were also measured during two waves of follow-up assessments, scheduled 6 and 18 months after the baseline assessments were completed. The imitation tasks were videotaped; coders later reviewed these videos and decided whether the children correctly imitated the target behaviors or not (i.e., pass, fail). Inter-observer reliability was established between two independent observers (ICC scores ranged between .97 to .98).
Results: Results revealed significant correlations between all four measures of imitation skills and children’s expressive/receptive language skills at follow up (waves 2 and 3). However once initial language skills and chronological age were statistically controlled, only some of these correlations remained significant. That is, partial correlations revealed significant correlations between children’s verbal imitation scores and their receptive language scores at wave 2 (partial-r=.486, p<0.05) and wave 3 (partial-r=.396 p<0.05); similarly, a significant partial correlation was found between object imitation and children’s expressive language skills at wave 3 (partial-r=.439, p<0.05). Finally, oral-motor imitation skills predicted both receptive (partial-r=.512, p<0.01) and expressive (partial-r=.393, p<0.05) language skills at wave 2.
Conclusions: This research reveals a complex pattern of relations between different domains of imitative behavior and children’s subsequent language gains. The ability to imitate spoken words appears to predict gains in a child’s ability to understand and respond to language, but not necessarily their ability to produce spoken language. The opposite is true for a child’s ability to imitate object manipulations. Only oral-motor imitation skills appeared to predict both expressive and receptive language gains.
Support: This research was supported by CPEA Grant HD-DCD35470; the M.I.N.D. Institute Research Program; an Autism SPEAKS Early Intervention & Treatment Grant.
Objectives: In this longitudinal study, we administered a comprehensive imitation battery that evaluates four types of imitative behaviors: manual, object, oral-motor and verbal (Rogers, Hepburn, Stackhouse and Wehner, 2003). We aimed to evaluate predictive relations between these four types of imitative behaviors and children’s subsequent gains in language skills.
Methods: The sample consisted of 34 individuals with autism (chronological age: mean = 55.9 months; SD = 11.9 months), who participated in the control condition of a randomized controlled intervention trial. At baseline, children were administered the imitation battery. A total of 24 imitation probes were administered over the course of 2 lab visits, with 5-7 tasks in each category. In addition, we evaluated children’s receptive and expressive language abilities using the Mullen Scale of Early Learning (Mullen, 1995). Children’s language abilities were also measured during two waves of follow-up assessments, scheduled 6 and 18 months after the baseline assessments were completed. The imitation tasks were videotaped; coders later reviewed these videos and decided whether the children correctly imitated the target behaviors or not (i.e., pass, fail). Inter-observer reliability was established between two independent observers (ICC scores ranged between .97 to .98).
Results: Results revealed significant correlations between all four measures of imitation skills and children’s expressive/receptive language skills at follow up (waves 2 and 3). However once initial language skills and chronological age were statistically controlled, only some of these correlations remained significant. That is, partial correlations revealed significant correlations between children’s verbal imitation scores and their receptive language scores at wave 2 (partial-r=.486, p<0.05) and wave 3 (partial-r=.396 p<0.05); similarly, a significant partial correlation was found between object imitation and children’s expressive language skills at wave 3 (partial-r=.439, p<0.05). Finally, oral-motor imitation skills predicted both receptive (partial-r=.512, p<0.01) and expressive (partial-r=.393, p<0.05) language skills at wave 2.
Conclusions: This research reveals a complex pattern of relations between different domains of imitative behavior and children’s subsequent language gains. The ability to imitate spoken words appears to predict gains in a child’s ability to understand and respond to language, but not necessarily their ability to produce spoken language. The opposite is true for a child’s ability to imitate object manipulations. Only oral-motor imitation skills appeared to predict both expressive and receptive language gains.
Support: This research was supported by CPEA Grant HD-DCD35470; the M.I.N.D. Institute Research Program; an Autism SPEAKS Early Intervention & Treatment Grant.