Simulation of other-initiated repair using AAC

Area of Science:

• Augmentative and Alternative Communication (AAC)
• Speech-Language Pathology
• Human-Computer Interaction

Background:

• Conversational repair is crucial for smooth communication.
• Speech-Output Technologies (SOTs) are used by individuals with communication impairments.
• The temporal demands of conversational repair are not well understood for SOT users.

Purpose of the Study:

• To assess the real-time capabilities of SOTs during conversational repair.
• To compare SOT utterance composition delays with temporal limits for oral communicators.
• To evaluate the communication performance of Augmentative and Alternative Communication (AAC) users in repair situations.

Main Methods:

• Simulated utterance composition delays for Other Initiated Repair (OIR) initiators using two SOT software applications.
• Analysis included conditions with and without word prediction.
• Comparison of simulated delays against a temporal limit derived from oral communication data.

Main Results:

• SOT utterance composition delays were substantially greater than the OIR limit, even at fast selection latencies.
• AAC production rates from literature review showed that most user groups could not produce full OIR utterances within conversational limits.
• Many users were unable to complete even a single selection within the identified temporal bounds.

Conclusions:

• Current SOTs and AAC technologies are not adequately timed for successful conversational repair.
• Findings highlight significant challenges for AAC users in participating in spontaneous repair sequences.
• Implications for designing future AAC devices to better support real-time conversational interaction.