Presented at ASSETS 2009
Joshua Hailpern, Karrie Karahalios, Laura DeThorne, Jim Halle
http://doi.acm.org/10.1145/1639642.1639675


Real-time computer feedback systems (CFS) have been shown toimpact the communication of neurologically typical individuals. Promising new research appears to suggest the same for the vocalization of low functioning children with Autistic Spectrum Disorder (ASD). The distinction between speech-like versus non-speech-like vocalizations has rarely, if ever, been addressed in the HCI community. This distinction is critical as we strive to most effectively and efficiently facilitate speech development in children with ASD, while simultaneously helping decrease vocalizations that do not facilitate positive social interactions. This paper provided an extension of Hailpern et al. in 2009 by examining the influence of a computerized feedback system on both the speech-like and non-speech-like vocalizations of five nonverbal children with ASD. Results were largely positive, in that some form of computerized feedback was able to differentially facilitate speech-like vocalizations relative to nonspeech-like vocalizations in 4 of the 5 children. The main contribution of this work is in highlighting the importance of distinguishing between speech-like versus nonspeech-like vocalizations in the design of feedback systems focused on facilitating speech in similar populations. (less info)