The Effectiveness of Joint Attention Intervention In Nonverbal Preschool Children with Autism

Background: Children with autism often demonstrate delays in both joint attention and play (Mundy, Sigman, & Kasari, 1990). Longitudinal studies have demonstrated that increased ability to initiate and respond to joint attention in preschool years has predicted language ability as many as eight years later (Siller & Sigman, 2002; Sigman & Ruskin, 1999). Furthermore, interventions that target joint attention skills can both improve both pretend play and expressive language, and children who may benefit the most from joint attention interventions are children who are nonverbal prior to intervention (Kasari et al., 2010).

Objectives: The current study will examine the effectiveness of a joint attention intervention for nonverbal children (less than 10 spontaneous words) in a preschool setting.

Methods: Fifteen nonverbal preschool children (ages 3-5) with autism were recruited from a non-public school to participate in the study.  Children had been enrolled for at last one year in 30 hours of intervention at the school based on Skinner’s verbal behavior approach.  Participants completed assessments on their cognitive skills (Mullen Scales of Early Learning) and language abilities (Reynell Developmental Language Scale).

        Children were randomized to either an observation-control (OBS) group (n = 8) or a joint attention treatment (TX) group (n =7). Children in the OBS group continued to receive their individualized education services through their school. Children in the TX group continued with their school services and were also pulled out from their classes twice a week for a 30-minute treatment session (total of 60 minutes of treatment per week) for 10 weeks. Treatment sessions were one-on-one sessions with an interventionist in a naturalistic play setting in which social communication gestures and functional play were targeted.

Children were observed in their classrooms for 20 minutes during their classroom free play at entry and at exit. Specially, children’s joint engagement and spontaneous social communication were observed in their classrooms.

Results: An ANOVA revealed that there were no significant statistical differences in mental age, F(1,13) = 2.721, p > .05, receptive language, F(1,13) = .508, p > .05, or expressive language, F(1, 13) = 1.122, p > .05 between the two groups at entry. With spontaneous gestures as the covariate at entry, an ANCOVA revealed that children in the TX group initiated more spontaneous gestures (joint attention and behavioral requests) than children in the OBS group at exit, F = 12.82, p < .01. With the percentage of time spent in unengaged state as the covariate at entry, an ANCOVA also revealed that children in the TX group were spending less time in unengaged state than children in the OBS group at exit, F = 11.98, p < .05.

Conclusions: Findings from the study suggest that the joint attention treatment was effective in improving engagement and spontaneous social communication in nonverbal preschool children with autism. Children who were in the TX group were able to generalize the skills they had learned in their pull out sessions to their classrooms. Future studies should continue to develop interventions that target these skills for this population.