ISCA Archive SLaTE 2023
ISCA Archive SLaTE 2023

Enhancing ASR-Based Educational Applications: Peer Evaluation of Non-Native Child Speech

Simone Wills, Cristian Tejedor-Garcia, Catia Cucchiarini, Helmer Strik

With recent advancements in automatic speech recognition (ASR), ASR-based educational applications have become increasingly viable. This paper presents a preliminary investigation into whether peer evaluations of the speech produced during the use of these applications, by primary school-aged children, is reliable and valid. Twenty-one Dutch primary school children assessed non-native read speech in terms of intelligibility, accuracy, and reading performance. The children's judgements were compared to those made by adult Dutch speakers, as well as to the performance of the Whisper ASR system. The children proved to be reliable raters with agreement levels en par with the adult group, with findings indicating that primary school-aged children can provide peer evaluation of speech suitable for enhancing the feedback provided by ASR-based language learning applications.