Pragmatic Language Competence of School-Aged Children with Autism in Peer Conversations

Background: These years more and more verbally fluent children with autism spectrum disorders (ASD) are able to adapt into the mainstream education and develop peer relationships in their school age. Researches show that these children may still present some residual deficits, and the pragmatic language competence is one of them. However, little is known about these subtle difference from a quantitative and detailed aspect. Most of the standard tests of pragmatic language (parent/teacher report scales or checklists and structured evaluation) find that these pragmatic deficits may be the ability of turn-taking and autistic-specific speech acts. Since the pragmatic impairments sometimes only present in real-life social interactions, we set up a natural environment and explore children’s pragmatic language competence during peer conversations between typically-developed (TD) children and these children with ASD, which provides a new perspective in detecting and evaluating pragmatic language ability.

Objectives: This study aims to compare the communication acts, conversational skills and autistic-specific speech acts of these mandarin-speaking children with ASD and TD children during peer talk.

Methods: 13 ASD (mean age 9.15±1.34, IQ: 92.38±13.15) and 13 age-matched and IQ-matched TD children (mean age:9.08±1.65, IQ: 98.62±11.63) participated in this study. We used a strategic card game to induce the peer talk between the children with ASD and his/her TD friend/cousin（age difference less than 2 years old). Then we coded and analyzed the 10 minutes’ conversation by CHILDES (Child Language Data Exchange System). The coding system we used in communication acts is Inventory of Communicative Acts-Abridged (INCA-A). The coding aspects of conversation skills includes turn-taking, initiation, continuation (reply and expand), discontinuity and repair. Furthermore, we defined the special autistic-style speech into five types according to the ASD linguistic symptoms. The Chi-Squared Test and t test are used to test the significance of the difference and the Pearson correlation coefficient was used to analyze the clinical significance of the pragmatic language competence.

Results: The results suggested that compared to their own TD peer, children with ASD showed significantly less sum of conversation continuation (ASD:37.46±12.96, TD:66.69±23.45, p=0.002), especially less expand in his/her turn (ASD:16.38±8.69, TD:40.54±23.44, p=0.02), less ratio of conversation continuation to initiation (ASD:0.89±0.41, TD:1.63±0.64, p=0.004), more autistic-style speech (sum: p=0.005, type: p=0.002) and more discontinuity(p=0.02). The difference in communication acts, sum of turns, and conversational repair were not significant. Strong correlation was found between total score of ADOS and some pragmatic indications, such as sum of effective utterance (r=-0.61, p=0.02), type of autistic-style speech (r=0.70, p=0.007) and sum of conversation initiation (r=0.63, p=0.02).

Conclusions: The results of the present study gave new evidence about the pragmatic language competence of children with high functioning ASD. They tend to express more actively in changing topics and initiating turns, but their performance of continuation of a certain topic, clarifying messages and adding more related information in their own speech turns was not as good as their TD peers, which cannot be reflected from their ADOS score and appear to be more difficult to improve even if they have adapted to mainstream education.