Effects of Dictation, Speech to Text, and Handwriting on the Written Composition of Elementary School English Language Learners
Abstract
Previous research has shown that both dictation and speech-to-text (STT) software can increase the quality of writing for native English speakers. The purpose of this study was to investigate the effect of these modalities on the written composition and cognitive load of elementary school English language learners (ELLs). In a within-subjects experimental design, 21 ELLs completed persuasive texts in 3 modalities: handwriting, dictation to a scribe (DS), and dictation to STT. Repeated measures analyses of variance showed that DS, compared to handwriting, resulted in significantly higher text quality, word count, and fluency and a lower error rate and required effort. STT, compared to handwriting, resulted in significantly higher holistic text quality and a lower error rate and effort. We examined student and text characteristics associated with text quality in each modality. Both modalities present young ELLs with options for written composition.