Examining the Use of a Gamified Transcription Task to Support Perceptual Learning of Dysarthric Speech

Purpose: The primary aim of the current study investigated whether a lexical transcription task during perceptual training with dysarthric speech improves listener perception of the degraded speech signal. A secondary aim investigated whether adding a gamified component to the training task would further increase intelligibility gains. Method: Following an intelligibility pretest using stimuli provided by a speaker with ataxic dysarthria, 59 participants completed a perceptual training phase dependent on their assigned condition (lexical transcription of healthy speech & feedback [control], lexical transcription of dysarthric speech & feedback [LT], or lexical transcription of dysarthric speech & gamified feedback [LTG]). Participants then completed a intelligibility posttest. Results: Analysis of intelligibility data at pretest and posttest indicated no significant differences between any of the three experimental conditions. Further fine-grained analysis revealed that all participants, regardless of condition, experienced a 10% intelligibility gain during the training phase and a second 10% intelligibility gain during the posttest phase. This suggests that participants used the training phase to learn how to complete the transcription task and used the posttest to learn about dysarthric speech. Gamification, as it was deployed in this study, did not have any discernable effect on intelligibility improvements. Conclusions: Future studies investigating perceptual learning of dysarthric speech may benefit from adaptions to the transcription task used in the current study in order to reduce the cognitive load experienced by the listeners. Improvements to the gamification element of the task may also be improved in order to increase motivation during the training phase