Children with autism have difficulties in vocal communication and social interaction in multiple contexts. Previous studies have shown that contingent imitation was effective in increasing non-verbal communication, such as eye contact, and proximity in non-verbal children with autism. However, it remained unclear whether contingent imitation could improve children's vocal communication and social interaction. This study aims to examine whether using contingent imitation training with one non-verbal child with autism could improve his/her occurrence rate of vocal responses, vocal imitation, production of meaningful words, and vocal turn taking. One non-verbal child with autism (CA＝34 months, DQ＝52) participated in this study. Before starting the experiment, the child had exhibited only non-functional speech such as /a/, /te/, and /da/. We used a time series AB design for this study. In both the baseline and post-phase stages, the child engaged in free play with the experimenter. The experimenter sat face-to-face with the child without speaking to him. Every sixty seconds, the experimenter modeled a speech such as "Gimme" or "Train." In the training phase, the experimenter modeled a speech and contingently imitated every vocal response emitted by the child. For example, immediately after the child said "da," the experimenter contingently imitated the child's vocal response "da." The results of the experiment showed that the occurrence rate of vocal response and vocal imitation increased during both the training and post-phase stages, and the child acquired meaningful words such as "Gimme" and "Train." Furthermore, the number of instances of vocal turn taking and the duration of social engagement between the experimenter and the child increased. Hence, the results indicated that contingent imitation was effective to promote both vocal communication and social interaction in non-verbal children with autism.

論文