Teacher Attitudes That Affect Implementation of Evidence-Based Practices in Self-Contained Autism Support Classrooms in Public Schools

Background: The significant lifelong impairments associated with autism spectrum disorder (ASD), combined with the growing number of children diagnosed with ASD, have created urgency in improving school-based quality of care. Few empirically supported interventions have been effectively implemented in schools, the primary setting in which children with ASD receive services. EBPs for children with autism are intensive and require substantial training to achieve successful implementation. Several studies indicate that with training, teachers and classroom staff can implement autism EBPs; however, there are varying levels of EBP use and implementation fidelity, the degree to which an EBP is used as designed (Proctor et al., 2011; Suhrheinrich et al., 2013; Pellecchia et al., 2015; Mandell et al., 2013). Effective implementation ensures that EBPs go from being merely a good idea to actually being put into practice in organizations. A critical factor in effective implementation is the context in which implementation takes places. The field of implementation science has a variety of perspectives and advancements related to implementation, especially at the individual level, which have not been fully considered by autism researchers.

Objectives: The purpose of this study was to examine the individual factors associated with the implementation of EBPs (i.e., discrete trial training, pivotal response training, and visual schedules) for children with ASD in self-contained classrooms in public elementary schools.

Methods: Participants included 67 autism support teachers from 44 public elementary schools. Participants completed ratings of attitudes of EBPs using the Evidence-Based Practices Attitudes Scale, a 15-item psychometrically sound measure that assesses four general attitudes toward adoption of EBPs: appeal, requirements, openness, and divergence (Aarons, 2004). Teachers and classroom staff completed self-report ratings of the frequency with which they deliver each of the EBPs as a proxy for implementation fidelity (e.g. dose).

Results: We computed means and standard deviations to examine attitudes about EBPs and fidelity of each EBP. Despite the nested structure of the data (classrooms nested within schools), multilevel models (MLMs) were not conducted, as there was one teacher per self-contained classroom, which did not allow for clusters needed for MLMs. Instead, regression analyses were used. In the unadjusted models, separate linear regressions were used to test for associations between attitudes about EBPs and fidelity of each EBP. In separate adjusted models, we entered all variables of interest as independent variables with a bivariate association significant at p <.2 and fidelity of each EBP as the dependent variable. Attitudes about EBPs were significantly associated with teachers' use of discrete trial training (p=.01) and visual schedules (p=.04) but not pivotal response training.

Conclusions: The results of this study provide an in-depth understanding of individual-level factors that influence the implementation of EBPs for children with ASD in public schools. These data suggest that individual attitudes about EBPs may be important to consider when evaluating successful implementation of some autism EBPs in elementary schools. Positive attitudes about EBP use may lead to increased use when EBPs are highly structured (discrete trial training and visual schedules) as opposed to naturalistic (pivotal response training).