Moderating Effects of Verbal IQ on Social Competence Intervention Outcomes

Background: Students with autism spectrum disorders (and similar social challenges) often struggle social interactions with both peers and adults. Despite increased efforts to identify effective interventions to address these social competence challenges, few intervention studies examine characteristics that may be associated with higher or lower treatment response. Research implicates the role of verbal abilities on a variety of social functioning indicators of youth with ASD or similar social/behavioral challenges.

Objectives: The objective of the current study is to examine the moderating effect of student verbal IQ (VIQ) on treatment response, as measured by teachers’ reports of social functioning, to a specified social competence curriculum.

Methods: This study utilizes data from a four-year cluster-randomized trial examining the efficacy of the Social Competence Intervention for Adolescents (SCI-A, n = 146) versus school-designated business as usual social programming (BAU, n = 128) for a range of middle-school students identified with social challenges (e.g., ASD, emotional behavior disorders, ADHD). The SCI-A curriculum is grounded in cognitive behavior intervention and includes specific social content and instructional strategies across all lessons. BAU settings varied considerably in their scope and delivery. Students across both conditions were primarily male (>84%) and mean full-scale IQ (FSIQ) scores in the average range. The majority of students had eligibility for special education services, with over 75% under the categories of autism or emotional/behavioral disturbance. For each student, one general education teacher completed the Social Responsiveness Scale, 2^nd edition (SRS-2) at pre and post intervention.

Results:  Controlling for relevant behavioral covariates and FSIQ, we conducted multilevel models testing moderating effects of VIQ and of VIQ/nonverbal IQ discrepancy on SCI-A’s efficacy in improving teacher’ reports of social outcomes. We examined the effect sizes simple slopes at lower (85) versus higher (115) levels of VIQ and at lower (-11) versus higher (+11) levels of VIQ to NVIQ discrepancy. Results indicated significant interactions between condition (SCI-A versus BAU) and students’ VIQ (see Table 1) and the VIQ to NVIQ discrepancy (see Table 2). SCI-A had more positive impact than BAU on students’ social awareness (d = .38), social communication (d = .50), and social motivation (d = .35) at lower levels of VIQ; these patterns were more pronounced when students had discrepantly lower VIQ than NVIQ (awareness d = .97, communication d = .97, motivation d = 1.34). For those at higher levels of VIQ and with discrepantly higher VIQ than NVIQ, SCI-A was less effective than BAU in improving students’ social outcomes (ds > .47).

Conclusions: SCI-A demonstrated small to moderate effects on social awareness, communication, and motivation to be social for students with lower verbal abilities. SCI-A’s consistency of concept language, cognitive strategies for social interpretation, and scaffolded opportunities to learn and practice skills may better address the needs and challenges of students who have lower verbal abilities than their higher ability peers. These findings have implications for understanding the role of VIQ in intervention response and determining what intervention features may be more impactful for students with lower versus higher verbal abilities.