Association for Behavior Analysis International

The identification of reinforcers is critical to maximize skill acquisition for individuals with autism. This symposium presents four studies detailing refinements of preference assessment procedures. In Study 1, single-operant preference assessments (SOA) effectively identified preferences (and subsequent reinforcer efficacy) for edibles, tangibles, and social stimuli for 4 individuals with autism. In Study 2, response restriction (RR) and brief response restriction (BRR) assessments were conducted with 12 individuals with autism. There was strong correlation between the RR and BRR results, but the BRR took 79% less time to complete. Study 3 evaluated whether edible, tangible, and social stimuli presented in a single-class MSWO were displaced when presented in a combined-class MSWO. Results indicated that displaced stimuli functioned as reinforcers for three children with autism. Study 4 evaluated the effects of lag schedules of reinforcement on variability in play material selection using a preference assessment format, with 5 children with autism. Lag schedules were found to be effective at increasing variability of selections for only one participant. In all studies, interobserver agreement (IOA) data were collected in a minimum of 30% of sessions, and was above 90% for all participants. These studies highlight the importance of continuing to refine preference assessment procedures.

There are many preference assessment methodologies available for identifying preference across a range of stimuli. To date, one study has used a single-operant assessment (SOA) to assess preference for social interactions. The purpose of the study was to extend previous research by evaluating a SOA to identify preference for edibles, tangibles, and social stimuli with four individuals with autism. Outcomes of the SOA were compared to outcomes from an MSWO. The SOA identified highly-preferred stimuli and correlations with the MSWO were found for some participants. The effects of top-ranked stimuli identified from SOA were then evaluated on a novel skill. Results support the use of a SOA to identify preferred edibles, tangibles, and social interactions and found that participants successfully acquired targets using stimuli identified from the SOA. Three of the four participants maintained the targets acquired and that the SOA identified an effective and efficient reinforcer. Interobserver agreement (IOA) data were collected in a minimum of 30% of sessions, and was above 90% for all participants. The procedures of the single-operant assessment were reported to be socially valid. Results support the use of an SOA to identify preference across a range of stimuli.

The response restriction (RR) preference assessment (Hanley et al., 2003) combines elements of existing preference assessments, utilizes free operant responding, and restricts items while reducing the likelihood of aberrant behavior that may occur due to denied access or insufficient duration to access reinforcers. However, the RR criteria to restrict items can be difficult to implement and the assessment can last multiple hours. In the current study, Phase 1 evaluated the validity and total duration of a brief response restriction (BRR) preference assessment. A RR, BRR, and 80-min free operant assessment were conducted twice each with 12 participants with autism. Validity results indicated a mean .69 correlation between the RR and BRR results, but the BRR took 79% less time to complete. Phase 2 evaluated that procedural integrity and implementer preference of the RR and BRR. We used enhanced training materials (Graff & Karsten, 2012) to train 10 special educators to conduct both a RR and BRR assessment. Procedural integrity scores were higher for the BRR, specifically for restriction criteria, and implementer surveys indicated preference for the BRR. Interobserver agreement (IOA) data were collected in a minimum of 30% of sessions and agreement was above 90% for all sessions

Preference assessments yield valuable clinical information important for skill acquisition and behavior reduction programs. Research has demonstrated that combining edibles and leisure items within a MSWO stimulus array can displace preference for leisure items among individuals diagnosed with profound intellectual disability. To date, no study has evaluated the potential displacement of edible, leisure, and social stimuli when combined in a single-class multiple stimulus without replacement (MSWO) preference assessment with children who have an autism spectrum disorder. The purpose of this study was to evaluate whether stimuli that are highly preferred when presented in a single-class MSWO array are displaced when presented in a combined-class MSWO array. The reinforcing effectiveness of displaced stimuli was evaluated in a single-operant reinforcer assessment. Results from this study indicate that stimuli that were displaced in rank order by two or more places when presented in a combined array MSWO functioned as reinforcers for three children with autism spectrum disorder. Interobserver agreement (IOA) data were collected in a minimum of 30% of sessions and agreement was above 90% for all sessions.

Repetitive behavior is a defining feature of autism-spectrum disorder (American Psychiatric Association, 2013). Children with autism tend to make invariable selections of play materials relative to their typically developing peers (Bancroft et al., in press). Lag schedules have been shown to increase behavioral variability (e.g., Cammilleri & Hanley, 2005), but they are not always effective alone (e.g., Napolitano et al., 2010). The purpose of the present study was to examine the effectiveness of a treatment hierarchy consisting of lag schedules combined with other techniques for increasing variable responding in play contexts for five children with autism aged 4 to 8 years. Across 4 activities, participants were instructed to choose an item from an array and prompted to engage in the activity associated with that item. Items were replaced between trials to allow opportunities for invariable selections. Experimental control was demonstrated using a multiple-baseline design across activities. Lag schedules alone were effective at increasing variability of selections for one participant, with generalization across activities. Variability of selections for the remaining participants did not increase until modifications were included. Interobserver agreement (IOA) data were collected in a minimum of 30% of sessions for all participants; IOA was above 90% for all participants