Association for Behavior Analysis International

Approximately 30% of children with Autism Spectrum Disorder do not develop speech. As a result, there may be benefits from interventions that teach the use speech-generating devices to establish verbal behavior repertoires. Numerous studies have shown the effectiveness of teaching the use of speech-generating devices in the context of basic manding. However, extension of this research is needed to evaluate procedures for teaching advanced verbal behavior for children who use SGDs. This symposium will present empirical data related to teaching verbal behavior to children using speech-generating devices. The first single case study evaluated procedures aimed at teaching two nonvocal children with autism spectrum disorder to use a speech-generating device to mand to their peer and engage in listener responding. The second study replicates the procedures from Frampton, Wymer, Hansen, & Shillingsburg (2016) to teach children who use speech-generating devices tacts of noun-verb combinations using matrix training. Wendy Machalicek will sever as the discussant.

Target audience are practitioners, students, and researchers in the field of applied behavior analysis who work with individuals with autism who have limited communication skills.

Children with autism spectrum disorder with limited speech are often taught to use speech-generating devices as an alternative communication mode. Intervention with speech-generating devices often begins by teaching the child to mand for, or request, preferred objects. To date, research on teaching children to produce mands for actions via a speech-generating device is limited. In the present study, we evaluated procedures aimed at teaching two nonvocal children with autism spectrum disorder to use a speech-generating device to mand to their peer and engage in listener responding. Naturalistic teaching procedures were used to teach both speaker and listener behavior. We also assessed for maintenance of the skill over time. A multiple baseline design across participants design was used to evaluate the effectiveness of this intervention. Both participants acquired the mands to peer and listener responding. These results suggest the feasibility to teaching mands to peers and listener behavior to children who communicate using speech-generating devices.

Approximately 30% of individuals diagnosed with autism fail to develop vocal communication. For these individuals, Augmentative and Alternative Communication (AAC) systems such as manual signs, picture exchange communication system (PECS), and speech generating devices (SGD) are often used. Numerous studies have shown the effectiveness of selection-based pictorial systems to promote functional communication. However, there is a dearth of research on strategies to teach advanced verbal behavior using SGDs. The current study presents a replication and extension of previous work conducted with vocal children teaching tacts of noun-verb combinations using matrix training (Frampton, Wymer, Hansen, & Shillingsburg, 2016). Three males diagnosed with autism were exposed to matrix training with mastered tacts of nouns (e.g., “elephant”) and verbs (e.g. “reading”). Two matrices were constructed (Matrix 1 and Generalization matrix), using mastered nouns and verbs. Following baseline of the matrices, diagonal targets within Matrix 1 were trained (e.g., “elephant reading”). Post-tests were conducted for the Generalization matrix followed by post-tests for Matrix 1. Two participants showed recombinative generalization with the Generalization matrix after training of diagonal targets in Matrix 1. For the third participant, correct responding with the Generalization matrix targets was observed after training with four different matrices (Matrix 1 – 4).