The Effects of a Parent-Mediated Joint Attention Intervention on Children's Play Skills One Year Later

Background: Autism intervention literature has shown that targeting a pivotal skill can lead to improvements in other areas (Koegel, Koegel, & Carter, 1999). For example, interventions teaching symbolic play behaviors have led to improvements in children’s social abilities (Kasari et al., 2006; Thorp, Stahmer, & Schreibman, 1995; Stahmer, 2005). However, teaching joint attention behaviors did not result in improved symbolic play, which suggests that symbolic play may require more direct teaching (Kasari et al., 2006). This is contrary to theoretical beliefs that symbolic play is acquired through shared meaning with another person (Werner & Kaplan, 1963; Vygotsky, 1978) and does not require direct teaching.

Objectives: This study will examine if children assigned to an experimental parent-mediated intervention targeting joint attention will display greater gains in play behaviors up to one year later when compared to a control group. Potential moderators of response to intervention will be examined.

Methods: Seventy children diagnosed with autism (chronological age: M = 57.1 months, SD = 12.3) participated in a randomized controlled trial with two treatment conditions. The experimental group received 12 in-home parent-mediated intervention sessions targeting children’s joint attention behaviors in the context of toy play interactions. The control group received four in-home advocacy sessions. A battery of assessments, including the Structured Play Assessment (Ungerer & Sigman, 1981), was administered and videotaped at pretreatment, post treatment, and 1 year follow-up periods.

Four independent raters coded the Structured Play Assessment for novel and spontaneous play behaviors. A composite play score was calculated by adding the frequency of functional (e.g., feeding baby with spoon), symbolic (e.g., making the doll slurp out of a cup), and sequenced (e.g, pouring into cup, then drinking out of cup) play behaviors. Excellent inter-rater reliability was established based on approximately 70% of the videotapes (ICC = .94-.97).

Results: No significant differences in pretreatment characteristics (e.g., CA, nonverbal mental age and language abilities) were found between experimental and control groups. No significant differences were found for gains in play behaviors between the experimental and control group at post-intervention, t(58) = -0.31, p =.76, or 1 year follow up, t(49) = -0.54, p = .60.

To examine a potential moderated effect, the sample was split at the median of the children’s initial nonverbal mental ages. For children with low initial nonverbal mental ages, results reveal greater gains in play behaviors at 1 year follow up for children in the experimental group (M = 4.67, SD = 5.61) compared to the control group (M = -0.50, SD = 8.97), t(24) = -1.72, p < .05, one-tailed. There was no significant improvement immediately after intervention, t(29) = -0.915, p = 0.37. Children with incomplete data had equivalent baseline characteristics when compared to children with complete data.

Conclusions: Findings from this study suggest play skills of children with autism can be improved by means of a parent-mediated intervention designed to promote their joint attention abilities.