Association for Behavior Analysis International

Restrictive interventions are used to prevent aggression by children and adolescents hospitalized due to severe behavior disorders and mental illness. Positive reinforcement and organizational behavior management systems were used to reduce restrictive interventions without increasing the frequency of aggression. A behavior analyst, nursing manger, and staff for 3 patient care units (38 patient beds) designed a positively-based behavioral system (CALM). During baseline, staff were provided with competency-based training on delivering praise and working with groups. After baseline, patients were provided with frequent praise for pro-social behavior and also received back-up reinforcers. Staff were provided with performance feedback. An ABC design was used. The performance monitoring component of the system was the intervention. Frequency of restrictive interventions, behavioral incidents, and percent correct for staff performance were dependent measures. A reduction in the number of restrictive interventions in comparison to baseline occurred with the implementation of monitoring and feedback. An even greater reduction in use of restrictive interventions followed increased monitoring and feedback (see Figure 1). The ratio of restrictive interventions to behavioral incidents also showed clear improvement (see Figure 2). That is, while the number of restrictive interventions decreased, the number of behavioral incidents did not increase. Thus, the system reduced restrictive interventions without increasing behavioral incidents. Anecdotal evidence also suggests that a positive change in the culture of the unit also occurred.

The quality of treatment has a significant impact on the effectiveness of behavioral treatment for those on the autism spectrum (LeBlanc, Ricciardi, & Luiselli, 2005). Paraprofessional staff do not have the extensive training, both scholarly and clinical, that Board Certified Behaviors Analysts possess. However, the training of these staff is paramount in order to implement useful behavioral procedures to a greater population. As a result, treatment integrity is a worthwhile measure to implement as it can dictate the level of effective treatment (Peter, Horner, & Wonderlich, 1982; Gongola, Barton, Gongola, Rosales, and Speece, 2014; Wheeler, Baggett, Fox, and Blevins 2006). Self-monitoring procedures have been an effective tool to increase the treatment integrity among behavioral therapists (LeBlanc et. al, 2005; Mowery, Miltenberger, Weil, 2010; Richman, Riordan, Reiss, Pyles, and Bailey, 1988). This study examined the reliability and utility of using a performance development plan (PDP) to dictate the goals on staff and employer self-monitored Performance Matrix to increase treatment integrity among its participants.

The purpose of this intervention was to heighten staff awareness and increase the number of learning opportunities for children with autism receiving early intervention applied behavior analysis services at an outpatient program in southwest Michigan. The behavior technicians (staff) were provided a goal for the number of learning opportunities to provide in a half hour interval, a sheet to record the total number of trials per half hour interval, and a frequency clicker to track their current trial rates. Baseline data indicated that the children were receiving less than the target goal of learning opportunities, per half hour interval. Thus, a group goal was set for 80% of the children receiving services to have a five percent increase in learning opportunities over baseline per half hour. If staff met their group goal of 80% of children with the five percent increase across four consecutive weeks, they were entered into a prize drawing for a five dollar gift card of their choice.

Task interspersal is a common technique demonstrated to facilitate learning in children with Autism Spectrum Disorder (Koegel et al., 2010; Chong & Carr, 2005; Kelly & Holloway, 2015). However, this technique can be difficult to implement, with some teachers attaining as low as 67% treatment integrity in controlled, internally valid studies. Errors in implementation can negatively impact outcomes for children (Worsdell et al, 2005). The present study examines the types of errors incurred, compares two methods of staff training as to how these impact treatment integrity, and assesses whether modified data sheets reduce implementation errors. Seventy-eight undergraduate participants were trained to implement task interspersal and then video recorded while conducting three task interspersal sessions. One third of participants received behavioral skills training (i.e., description, video modeling, live modeling, and role play with feedback), 1/3 received an abbreviated training (i.e., role play with feedback), and 1/3 received the abbreviated training with modified data sheets. Initial analysis for 10 participants (complete analysis will be completed before the conference) revealed task interspersal schedule, material layout, and reinforcement to be the most frequent errors (Table 1). Findings will inform methods to improve efficiency of training staff to minimize implementation errors.

The present study seeks to extend the current literature on utilizing behavioral skills training (BST) to teach practitioners how to implement functional analyses (FA). Using a multiple baseline design across participants, this study will measure the effects of using a remote BST package on accurate implementation of FA procedures. Specifically, researchers will use the latest HIPAA-secure teleconsultation technology and BST when training practitioners who have no previous training in FA methodologies. Each participant will experience four phases, which include baseline (only instructions with simulated clients), BST (instructions, modeling, rehearsal, and feedback with simulated clients), post training probes (probes with simulated clients), and generalization probes (probes with actual clients). We will evaluate the effects of a remote BST package on levels of performance in simulated conditions as well as the effects of the remote BST on performance in generalization probes with actual clients.