Cette étude longitudinale porte sur le développement de la communication chez des enfants de 3 à 6 ans avec TSA, scolarisés en unité d’enseignement en maternelle. Elle s’appuie sur l’observation directe de situations filmées en classe, soumises à une micro-analyse de la communication et des interactions sociales à l’aide d’une grille structurée. La fréquence et la qualité fonctionnelle des échanges sont comparées dans les interactions avec les adultes et avec leurs pairs, en situation structurée (travail à table) et en situation non structurée (jeu libre), et dans leur évolution au cours du temps (première année vs dernière année de maternelle). Les résultats montrent que l’engagement socio-communicatif des enfants est généralement plus fréquent dans les activités structurées et avec les adultes, essentiellement en modes de réponses aux sollicitations. Cependant, les initiatives des enfants augmentent dans le temps en situation de jeu libre. Les interactions à but instrumental demeurent la majorité des actes communicatifs. Les buts de partage social sont davantage favorisés par la situation de jeu, où leur proportion augmente au cours du temps. La part des intentions de partage d’états mentaux évolue peu au cours des trois années. Les trajectoires individuelles de progression diffèrent selon les caractéristiques des enfants. Ceux qui présentent les scores les plus avancés dans les échelles développementales au début de la scolarisation montrent une progression plus importante des interactions sociales et de la communication en classe. Les données sont discutées afin d’optimiser les dispositifs scolaires inclusifs.

Inclusive preschool classrooms have been recently created (since 2014) in France for children with Autism Spectrum Disorder, giving them access to regular schools while adjusting to their specific needs. In these preschool small units, educative and mental health staff work together to support children's communicative and learning skills in order to help them to progressively attend regular classes in the same school. International research has reported the benefits of inclusive education on children with ASD, but studies mainly concerned children in full inclusion in regular classrooms, whose constraints, even with adapted programs, may not be suitable for all of them, especially young children. Moreover, studies conducted in specialized as well as in regular classrooms showed that the frequency and functional quality of the children's communicative exchanges may greatly vary, depending on the interlocutor and on the requirements of school situations (structured or unstructured). In this paper, we present a longitudinal study about communication development from age three to six in children with ASD attending an inclusive preschool program. On the basis of direct observations of situations filmed in class, variations of social communication were systematically examined as a function of partners (adults vs. peers), degree of structuration (work vs. free play), and over time.

Participants were seven children with ASD, aged 3 at the beginning of the study and attending the same preschool inclusive unit. They were regularly videotaped (twice a year during three years) in two types of classroom situations: structured (working) and unstructured (free play), interacting with adults and peers. Data reported here are for the comparison between the first and the last collection point (three years vs. six years). Videos were coded using a structured microanalysis grid to capture the quality (forms and functions) of social and communicative behaviors of children and of their partners. In parallel, standardized evaluations were conducted on children's clinical and developmental characteristics.

Data analysis showed a differentiated progression in quantity and quality of communication observed in preschool inclusive classrooms according to contextual parameters. Children's social-communicative engagement was generally more frequent in structured than unstructured activities, but initiations progressed over time in unstructured activities. Children interacted much more often with adults than with peers, with no significant increase in this gap over the years. The quality of children's communication improved both in forms and functions, but to varying degrees depending on the context. During structured activities, children became more able to use multimodal coordination (e.g., verbal/nonverbal) within the same communicative act while during unstructured activities they more often use one modality communicative acts. Concerning the purpose of interactions, children most often communicate for instrumental purposes than for social sharing purposes, a difference that persisted over time. A single subject analysis revealed important variations in pace and extent of developmental change between three and six years, with an increased progression observed in children with initial higher social and cognitive developmental levels.

Findings are discussed in the perspective of optimizing the implementation of inclusive preschool educative settings for children with ASD. In particular, three points seem to deserve reinforced support: (a) to promote opportunities for children's initiatives in addition to soliciting responses; (b) to enhance the opportunities for social sharing; (c) to support peer-to-peer interactions during play or work co-activities.