Association for Behavior Analysis International

Although much has been written about prompts and their role in teaching learners with autism, the literature on prompting procedures is mixed and limited. Important questions such as, when response prompts may be more effective than stimulus prompts; or does the timing of prompts matter; or whether prompts following errors function as reinforcers, to name but a few, await more empirical answers. This symposium will present four applied research examinations of questions about prompting procedures conducted at two different day-school programs for children with autism each using an applied behavior analytic approach to instruction. The studies compared stimulus and response prompts, two different prompts used in error correction, an errorless prompting procedure with prompts following errors, and the effects of reducing prompts after errors and increasing token delivery on a variety of skill acquisition programs. The results add to behavior analytic literature on prompting procedures and provoke further analysis of prompting variables in the role of teaching people with autism.

Throughout the applied behavior analytic literature, response prompts such as verbal instructions, modeling, and physical guidance have been critical components of intervention packages that have effectively established a breadth of skills with individuals with autism. Although the use of response prompts is well documented, the descriptions of the procedures for implementing such prompts varies significantly and the timing of those prompts (i.e., before or after an error) is often unclear. The purpose of the present study was to examine the effects of a training protocol on: (a) the rate of tokens and prompts delivered by the instructor, (b) the percentage of prompts delivered before and following errors, and evaluate their effects on the participants on-task behavior. The training protocol was examined using a multiple probe across instructors design that was replicated across two students. The results showed that the implementation of the training protocol produced systematic increases to the participants' on-tasks behavior. Interobserver agreement was collected on all dependent variables across 50% of sessions, data ranged from 60% to 100% with a means ranging from 84% to 89%.

Literature has shown effectiveness of treatment packages that included activity schedules and manual prompts with graduated guidance to teach individuals with autism to independently complete a variety of skills. Throughout the literature, however, there is variability in the descriptions of how to implement and fade manual prompts using graduated guidance. The purpose of this study is to evaluate the effects of implementing graduated guidance as: 1) a systematic fading procedure and 2) a consequence that was only delivered if the participants made an error. The effects of the prompting procedures were examined using an alternating-treatments design embedded within a multiple-probe across-participants design to increase participants' use of an activity schedule. The results showed that the use of graduated guidance with a systematic fading procedure produced more criterion performances across the acquisition, generalization and maintenance conditions when compared to graduated guidance delivered only if the participants made an error. Stimulus generalization was programmed for and demonstrated across materials and in the absence of direct adult supervision. Overall, the results contribute to the literature by suggesting superiority of systematic prompt fading strategies when compared to prompting in response to errors. Additional implications of this finding are also discussed.

Evaluating the relative efficacy and efficiency of established teaching methods across skills and learners is critical. While the use of both response prompting and stimulus prompting are common and effective methods of developing new skills in Discrete Trial Teaching for children with autism, there may be advantages, disadvantages and indications for each method. The current study compares the use of stimulus prompts (pointing) and response prompts (manual guidance) for teaching young children with autism to make new auditory-visual discriminations. Both methods utilized most-to-least prompt fading. Three out of four participants acquired items with both methods. The fourth participant only acquired the skill with stimulus prompting. Results suggest that learners' responses to both types of prompting should be empirically evaluated, and given comparable effectiveness, stimulus prompting should be considered less intrusive. Results are discussed in terms of other factors related to practitioners' choice of prompt type, prompt effectiveness, probability of errors and complexity of implementation.

Discrete Trial Training (DTT) is an important component of intervention and instruction in Applied Behavior Analysis (ABA) and is extensively used to teach individuals with autism and other developmental disabilities new behaviors and new discriminations among stimuli. Errorless teaching is an instructional strategy that is often used in DTT in order to reduce errors from occurring. However, because errors will inevitably occur during teaching, empirically testing and validation of error correction techniques used in DTT is important for the development of best clinical practices for DTT and ABA instruction. Therefore, in the current paper, we empirically compare two error correction procedures for teaching receptive identification in four learners with autism spectrum disorder: a follow-up trial using stimulus prompting and the use of corrective information (without the requirement of a follow-up response). We found that the more effective error correction strategy differed across learners. Implications for reducing prompt dependency and for individualizing therapies across learners with autism spectrum disorder and developmental disabilities will be discussed.