Association for Behavior Analysis International

Treatment integrity, also known as procedural fidelity, is the degree to which intervention steps are implemented with accuracy. Behavioral skills training (i.e., instructions, modeling, rehearsal, and feedback) is an effective procedure used to train staff across a range of repertoires (e.g., implementing skill acquisition programs, verbal behavior programs). The purpose of this symposium is to highlight four studies evaluating training procedures to implement interventions with high levels of integrity. The first presentation will share findings following the training of three parents to implement variations of discrete-trial instruction (DTI) with their children with autism and parent preference for DTI variations. The second presentation will describe how the authors identified variables impacting special education teachers selection of academic interventions and discuss variables impacting integrity in classroom settings. The third presentation will describe the use of video modeling with voiceover instruction when training teachers to implement token economies with children with autism. Finally, the fourth presentation will describe the devleopment of a cost-effective pre-service training package to teach three animal shelter volunteers to implement a dog walking and enrichment protocol (DWEP) through a video training package.

Special education teachers may select interventions for their students for a variety of reasons (e.g., familiarity with an intervention) that are not yet well understood. Once the teacher selects an intervention, it remains unclear whether the teacher and classroom staff are implementing the intervention as it is described in the literature. The purpose of the investigation was to identify variables impacting special education teachers selection of academic interventions and examine whether teachers implemented the intervention with integrity. We collected survey data and conducted observations of conditional discrimination training procedures in special education classrooms in Oregon. We examined the level of integrity that classroom staff implemented trial-based instruction, with most observations including instruction delivered by instructional assistants instead of special education teachers. Results indicated that educators implemented 50% of the components of trial-based instruction with integrity at or above 80%. Variables such as the mastery level of the task being presented impacted levels of integrity. We will further discuss variables impacting integrity in classroom settings and provide recommendations for future research and practice.

The use of token economies is frequently recommended in early intervention manuals and is reported to be commonplace in clinical practice. However, limited research is available to help guide clinicians in how to best train staff to implement token economies. To help address this void, the present study evaluated video modeling with voiceover instruction to train four staff trainees to implement a token economy. Initially, we evaluated the staff trainees integrity with a simulated consumer (i.e., an adult acting as a child). Generalization was programmed for and assessed with an actual consumer (i.e., a child with autism). The results demonstrated that video modeling was an effective approach to training. Staff trainees also demonstrated high levels of integrity up to 2-months following the completion of training. Multiple measures of validity were also completed and provide evidence for the content validity of the training video and the social validity of the goals, procedures, and outcomes. Together, these results support the usefulness of video training and suggest that performance feedback may not be a necessary component of training. We will discuss these findings in light of previous research and provide suggestions for future research.

Volunteers are ubiquitous to non-profit service organizations, yet methods to efficiently and cost-effectively train volunteers are relatively underexplored in the literature. The current study aimed to develop a cost-effective pre-service training package to teach three animal shelter volunteers to implement a dog walking and enrichment protocol (DWEP). Following the shelter's traditional live training, volunteers implemented just over half of all DWEP steps correctly (M = 55.2%). DWEP integrity improved when participants completed a video-based self-training package (M = 75.3%), but did not reach the pre-established mastery criterion of 85% fidelity with zero safety errors. During coaching, which consisted of modeling and positive and corrective feedback, integrity improved (M = 90.6%), yet only two of three participants met criterion performance. High integrity performance was observed for two of three participants at 1- and 4-week follow-up observations. Though creation of the video-based training package used in the study required approximately 13.25 hours longer than preparation of live training, live training required between 30-50 minutes with a shelter staff member with wide variability and safety of content observed. When used in place of life training, the video training package could save money for the organization in as few as 13 volunteer training sessions.