Tracking Changes in Social Skills: Weekly Behavioral Coding During a Social Skills Intervention Group

Background:

            Social skills interventions for children with Autism Spectrum Disorder (ASD) are often evaluated through child and parent report questionnaires administered pre- and post-intervention. Few studies have used behavioral coding during group time to evaluate the efficacy of social skills interventions.  

Objectives:

            The aim of this study was to determine whether children with ASD enrolled in a social skills intervention altered the frequency of their vocalizations and social interactions during group time.

Methods:

            Fourteen children with ASD, ages 10-16, participated in the current study. Participants attended a clinical social skills intervention for 1.5 hours weekly over 22 weeks. The intervention curriculum was a modified version of the Social Adjustment Enhancement Curriculum (Solomon, Goodlin-Jones, & Anders, 2004). Participants’ vocalizations and social interactions were coded weekly during an unstructured game time that involved playing board games and cards with other intervention participants.

            Participants’ vocalizations were coded as Initiating, Responding, or Other. Vocalizations directed toward another person were coded as “Initiating” in the absence of a conversation and “Responding” in the presence of a conversation. Vocalizations that were not directed toward another person (e.g., self-talk) were coded as “Other”. Participants’ social interactions were coded by interaction partner: interaction with one peer, interaction with one leader, interaction with a group of peers, interaction with a group of peers and leader(s), or by self (i.e., no interaction partner).

            Behaviors were coded in 20-second intervals by a team of trained undergraduate students; inter-rater reliability for the coded variables ranged from acceptable to excellent (alpha = 0.74 – 0.96). Two-level HLM models were used to analyze the data, with weekly behavioral coding scores nested within persons. HLM analyses were run separately for each coded variable, and age, verbal IQ, and gender were evaluated as potential predictors of each coded variable.

Results:

            From the beginning to the end of the intervention, participants made fewer Initiating vocalizations, t(265) = -5.48, p < 0.01, fewer Other vocalizations, t(265) = -3.31, p < 0.01, more Responding vocalizations, t(265) = 2.06, p = 0.04, and spent more time interacting with a group of peers, t(264) = 2.47, p = 0.01. Younger participants showed a steeper increase in the amount of time spent interacting with a group of peers than older participants, t(264) = -2.73, p = 0.01.

Conclusions:

            The observed decrease in Initiating vocalizations, although initially surprising, is coupled with an increase in Responding vocalizations; this finding suggests that participants were more frequently engaged in conversation, potentially resulting in fewer opportunities to initiate new conversations. This increase in conversation and peer interactions suggests the development of more stable interaction groups, or perhaps even friendships. These results demonstrate behavioral changes in interaction dynamics over the course of a social skills intervention; future research should utilize a control group that does not receive the intervention in order to determine the degree to which behavioral changes are driven by the intervention curriculum versus repeated interactions with peers.