Co-Principal Investigators: Odom, Samuel; Kraemer, Bonnie; Hall, Laura J.; Dickson, Kelsey

Purpose: The purpose of this efficacy study is to evaluate the comparative efficacy of two interventions for middle school students on the autism spectrum. Students with autism have challenges related to social skills and executive functioning that impact educational outcomes, yet there is limited evidence about effective school-based programs addressing these core areas of need for pre-adolescents with autism. The proposed study will address this need by evaluating an intervention that targets social skills, the Program for Education and Enrichment of Relational Skills (PEERS), compared to an intervention that targets executive functioning skills, Unstuck and On Target (UOT), on student social skills and executive function. Both of these programs have generated positive outcomes in initial tests of effects by the program developers. However, neither have been examined in a cluster randomized control trial in public middle school settings.

Project Activities: The project will use a three-arm cluster randomized control trial to compare PEERS, UOT, and a business-as-usual control condition. Schools will be randomly assigned to one of the three groups. Approximately 468 students with autism across four cohorts will receive the intervention during the spring semester. Assessment timepoints are the end of first semester (pretest), the second semester (posttest), and during first semester of the following school year (follow-up). The assessments include direct assessments and observations to assess the specific skills targeted in each intervention as well as student social skills, executive functioning, and learning barriers and facilitators.

Products:  This project will result in evidence of the comparative efficacy of PEERS and UOT for middle school students with autism on student social skills, executive functioning and learning facilitators and barriers, as well as information about the cost and cost-effectiveness of the programs. The project will also result in a final dataset to be shared, peer-reviewed publications and presentations, and additional dissemination products that reach education stakeholders, such as practitioners and policymakers.

Structured Abstract

Setting: The research will take place in 78 middle schools in North Carolina and Southern California.

Population/Sample: The team will recruit between 4 and 8 middle school students per school (approximately 468 students) with an educational label of autism and an IQ of 70 or higher.

Intervention: The PEERS intervention targets social skills related to areas such as conversation, humor, friendships, and handling conflict. Skills are addressed through weekly lessons that include homework review, didactic lessons, role play, and behavior rehearsal. Additionally, parents of PEERS participants will receive two parent training sessions delivered by the PEERS coach during the semester. The UOT intervention targets executive function skills, such as flexibility, compromising, planning, goal setting. Skills are addressed through weekly lessons delivered in a group format, classroom activities, homework, and two parent sessions. UOT focuses on enhancing flexibility in thinking, problem solving, and planning. The interventions will be each be implemented over two 45-minute sessions per week across 16 weeks and will be implemented by school-based staff.

Research Design and Methods: The project will use a three-arm cluster randomized control trial to compare PEERS, UOT, and a business-as-usual control condition. Schools will be randomly assigned to one of the three groups (26 schools per group). Approximately 468 students with autism across four cohorts will receive the intervention during the spring semester. Assessment timepoints are the end of first semester (pretest), the second semester (posttest), and during first semester of the following school year (follow-up).

Control Condition: Students in schools randomized to the control condition will receive the typical programming provided to students related to social skills and executive functioning

Key Measures: The research team will collect assessments from students and teachers to evaluate student outcomes before, immediately after, and then at a 6-month follow-up after the intervention. The proximal outcomes are program specific; a student questionnaire and observation will be used to examine targeted social skills from PEERS, and direct student assessment and observation will be used to examine targeted executive functioning skills from UOT. The distal measures examine generalized performance of social skills (Social Responsiveness Scale-2) and executive functioning (Behavior Rating Inventory of Executive Function-2) using teacher questionnaires. Assessment of learning facilitators will be collected using student and teacher questionnaires (Academic Competence Evaluation Scale — Academic Enablers subscale). Assessment of behaviors that may serve as learning barriers include interfering behaviors (Sutter-Eyberg Student Behavior Inventory-Revised), and anxiety (Anxiety Scale for Children — Autism Spectrum Disorder). Additionally, data on elements related to implementation that have practical implications (i.e., social validity, fidelity of implementation, cost) for the schools in all three groups will be collected.

Data Analytic Strategy: The primary analyses will use multilevel, hierarchical linear models (HLM) to account for the nesting of students within schools to examine differences between the three groups. To examine student-level moderation questions, the HLM will be modified to include interactions. The team will also examine practical factors related to implementation, such as fidelity of intervention and social validity using descriptive statistics.

Cost Analysis: The research team will use the ingredients method to calculate the costs of each active intervention and the control groups. They will use the cost analysis and effect sizes on distal student outcomes generated from the impact analysis for the cost-effectiveness analyses.