Fidelity of Implementation of Evidence-Based Practice in Community Classrooms

Background: Several behavioral methods have been identified as evidence-based practices (EBP) for children with autism spectrum disorders (ASD). In research studies examining the efficacy of these interventions, interventionists are required to meet strict fidelity of implementation criteria. It is likely that an integral part of moving these EBP into community settings is the correct implementation of these interventions by community providers. However, these EBPs require extensive training and may lack the flexibility needed for application in schools with more limited resources. Teachers express pervasive skepticism about the clinical utility of evidence-based ASD practices due to difficulties with adaptation to clinical settings (Stahmer et al., 2005). To date, fidelity of implementation of these interventions in classroom settings has not been tested. In addition, factors related to teachers’ ability to master the intervention strategies (e.g., attitudes toward EBP, teaching experience) have not been examined. Understanding areas of strength and difficulty for teachers in using these interventions and factors related to appropriate implementation is essential for translating EBPs to school programs. The Strategies for Teaching based on Autism Research (STAR) program is a comprehensive curriculum that uses EBPs to teach academic and social skills. Strategies include the highly structured discrete trial teaching (DTT), and a more naturalistic behavioral strategy pivotal response training (PRT) both of which are typically conducted individually. STAR was developed specifically for classrooms and is starting to be widely disseminated; however treatment integrity in community settings has not been systematically examined.

Objectives: To examine fidelity of implementation of the STAR program components, including DTT, PRT, and classroom environmental and visual strategies in urban classrooms.

Methods: Teachers in 20 classrooms assigned to the STAR condition of a randomized controlled field trial were videotaped monthly using DTT and PRT during the school day. In addition, classroom environments were coded for specific STAR program elements. Videotapes were coded for fidelity of implementation and rated for overall quality of teaching by undergraduate research assistants blind to the study hypotheses. Reliability of data coding was greater than 80% on all measures.

Results: 100% of teachers were willing to implement the DTT method on classroom taping days. Only 50% of teachers were willing to implement PRT on taping days.  Of those that were videotaped using each method, 40% met overall fidelity of implementation in DTT and 0% in PRT. Quality of teaching ratings and use of classroom environment strategies were higher for teachers who met fidelity of implementation in DTT. 

Conclusions: Urban public school teachers can implement evidence based intervention strategies. Teachers more frequently used antecedent strategies correctly than consequences in both DTT and PRT. Relationships between teacher characteristics and fidelity, and implications for community-based training and implementation will be discussed.