The effect of dyslexia on the production of speech errors, repairs and disfluencies

Abstract

Tests of verbal working memory are used to measure reading skills in individuals with dyslexia (Shaywitz et al. 2005; Ogino et al. 2017). The production of clear and fluent speech is also dependent on verbal working memory due to a limited capacity to retain and rehearse information and therefore less opportunity to check for errors prior to speech production (Jacquemot and Scott 2006). Despite these links, there is limited evidence regarding the association between dyslexia and the production of speech errors, repairs and disfluencies. The primary aim of the present study is to investigate the link between dyslexia and speech error production, while considering the effect of working memory capacity on speech errors, repairs and disfluencies. An additional aim of the study is to explore the effect of gender on working memory and speech errors. The working memory capacity of each participant is measured using three working memory tasks: reverse digit span, reorganising digits and non-word repetition. Speech errors are induced experimentally using a tongue-twister paradigm while speech repairs and disfluencies are naturally elicited through a spontaneous speech sample. The results are analysed using statistical analysis, exploring the relationship between working memory and speech errors, repairs and disfluencies, and investigating the difference between the speech errors produced in the dyslexia and control groups. It is concluded that individuals with dyslexia produce more speech errors, repairs and disfluencies than those without dyslexia, and that this is linked to verbal working memory. This contributes to evidence that tests of speech errors, repairs and disfluencies could be used as a predictor of dyslexia. The significance of these results, however, is dependent on the type of speech error elicitation task used.