The Social Infrastructure of Autism Treatments in Schools

Background:  Most children with autism spectrum disorders receive intensive educational interventions as their primary form of treatment. These treatments often require collaboration among school staff and between home and school, because children diagnosed with autism have difficulty transferring skills learned in one setting to another. This study investigated the socio-economic and organizational barriers that impacted the alignment of interventions for children diagnosed with ASD, both among school staff and between home and school interventions. 

Objectives:  This research study used innovative network techniques to develop a new approach to tracking 1) the delivery of autism interventions in differently configured schools and 2) the alignment of autism interventions across school and home settings. We provide two network maps that illustrate different school level configurations of autism interventions and 45 individual child network maps the illustrate the degree of overlap among interventions provided by school staff members, family members and external providers in each of the two school settings. While the ultimate goal is to design interventions that facilitate intervention alignment, for this study, our objective was to create a baseline measure of the degree of variance in intervention alignment among individual children situated in the same school, across socio-economic categories.

Methods:  We conducted 200 hours of field observations that generated 600 pages of digital field notes, 31 semi-structured transcribed interviews and 90 network surveys with 90 school teachers, aides, clinicians and parents, about forty five children diagnosed with autism spectrum disorder, who were situated in two differently configured public schools in a large metropolitan area. The study employed a “saturation sample”, which included recruitment of all parents of autistic children and all staff that provided interventions for those children at the two field sites. Interventions were rated as established, emergent or un-established using the National Standards Report (20009). Collection and analysis of qualitative data was done in tandem, allowing the development of working hypotheses, the adjustment of data collection strategies, and the pursuit of new data regarding interactions discussed in interviews and during observations. Coding categories were developed from theoretical review and analysis of interview and observational data. All interviews and observation notes were read several times, using memos, pattern coding, case analysis (across social class), and written summaries to aid in assuring the quality of data collection and provide opportunities for on-going data analysis. We used Dedoose to analyze the qualitative data. Parent and autism provider network survey data will be analyzed using ORA for dynamic, multi-modal networks and network visualization.

Results:  Two school configurations were identified:  1) a “school-within-a-school” model and 2) a maximum inclusion model. Intervention network maps for low income children had fewer interventions rated at established or emergent, with lower levels of alignment in the “school-within-a-school”, compared with the maximum inclusion model. 

Conclusions:  Preliminary findings indicate that school design may shape alignment of individual interventions for children diagnosed with ASD.