Association for Behavior Analysis International

While research has clearly demonstrated the efficacy of stimulus preference assessments in the identification of potential reinforcers, practitioners still rely primarily on indirect methods to select these stimuli (Graff & Karsten, 2014). In the current symposium, we will explore and discuss procedural and theoretical parameters of direct preference assessment in the attempt to increase audience competence in the use of direct methods. Basic procedural issues, as well as complex questions regarding potential sources of stimulus control within the assessment context will be discussed. The focus of this symposium is not only stimulate future research, but to also challenge practitioners to increase their use of direct over indirect assessment methods.

This presentation will review three recent studies that examined the predictive validity of brief MSWOs conducted in electronic pictorial and video formats with children with autism. The first study will demonstrate the predictive validity of electronic pictorial preference assessment, conducted on an iPad, in predicting reinforcing efficacy of toys. The second study is a replication of the first study, except stimuli were depicted in video format. Finally, the third study examined the extent to which a video-based preference assessment without contingent access to chosen activities (e.g., jumping on a trampoline or playing video-games) corresponded to a video-based preference assessment with contingent access. We also examined the reliability of instructor reports in predicting obtained student preference for activities. All three studies demonstrated the effectiveness of these novel formats of preference assessments and further highlight the importance of systematic evaluation of student preference in early-intensive behavioral intervention settings.

Several research studies have demonstrated the success of using preferences as reinforcement for children with autism and other developmental disabilities (Roane, et. al., 1998). However, children with autism often become obsessed with just one or small set of items, which can relate to higher levels of challenging behaviors (Mancil, 2009). Thus, the purpose of this study was to examine the effects of pairing procedures between preferred items and novelty items and the effects on subsequent item engagement and challenging behaviors during operant play conditions. A multi-element design was used to compare engagement time and challenging behaviors between highly preferred items and novelty items. Data was collected via iPad during 5-minute sessions. Prior to pairing procedures, preference assessments were conducted keeping response effort levels equal across items to identify the highly preferred items. A multiple stimulus with replacement and a paired stimulus preference was conducted on each participant. In addition, novelty items were identified for each participant and tested to ensure a zero level of engagement prior to pairing Novelty items were paired with highly preferred items for each participant. Pairing procedures consisted of requiring the participant to engage with the novelty item with the highly preferred item simultaneously. Results indicate that participants engaged in play with novelty items for significantly higher periods of time and challenging behaviors decreased following pairing procedures with highly preferred items. This study potentially impacts planning for individuals with limited preferences and obsessive interests. IOA was 95% across all conditions and reliability with each observers data was 100% across all observers.

Considerations in selecting items to include in a stimulus array for preference and subsequent reinforcer assessments are seldom addressed in the literature. While stimulus selection in preference assessments works on a utilitarian level, there is evidence to suggest that researchers might need to take additional precautions for stimulus selection practices. Specifically, research that does not consider the difference in magnitude between and within items in a stimulus array may result in a lack of correspondence between preference level and reinforcer efficacy (Halbur, Linn, & Witts, 2015). In their study, Halbur et al. (2015) found that preference levels of varying portion sizes of pizza did not reliably predict reinforcer efficacy in subsequent progressive ratio reinforcer assessments. It is our goal to extend DeLeon, Frank, Gregory, and Allman (2009), which demonstrated a correspondence between paired-stimulus preference assessments and progressive ratio reinforcer assessments, by considering magnitude across preferred and non-preferred items in a stimulus array. We conclude with a discussion on practical limitations, additional research considerations, and future directions in stimulus array selection.