Text-to-speech: what the research says

Craig, S. D., & Schroeder, N. L. (2019). Text-to-Speech Software and Learning: Investigating the Relevancy of the Voice Effect. Journal of Educational Computing Research, 57(6), 1534–1548.

Technology advances quickly in today’s society. This is particularly true in regard to instructional multimedia. One increasingly important aspect of instructional multimedia design is determining the type of voice that will provide the narration; however, research in the area is dated and limited in scope. Using a randomized pretest–posttest design, we examined the efficacy of learning from an instructional animation where narration was provided by an older text-to-speech engine, a modern text-to-speech engine, or a recorded human voice. In most respects, those who learned from the modern text-to-speech engine were not statistically different in regard to their perceptions, learning outcomes, or cognitive efficiency measures compared with those who learned from the recorded human voice. Our results imply that software technologies may have reached a point where they can credibly and effectively deliver the narration for multimedia learning environments.

Douglas, K. H.; Ayres, K. M.; Langone, J.; Bell, V.; & Meade, C. (2009). Expanding literacy for learners with intellectual disabilities: The role of supported etext. Journal of Special Education Technology 24(3), 35-44.

This quasi-experimental research study, published in a peer-reviewed academic journal, investigated the effects of presentational, translational, illustrative, instructional, and summarizing supports on the reading and listening comprehension of students with moderate intellectual disabilities through a series of single subject experiments. For the purpose of this annotated bibliography only the text-to-speech results will be summarized. The participants included eight students: five attended high school, and three attended middle school. Short reading passages at the fourth grade reading level were imported into Microsoft Reader, software that provides both text-to-speech and word-by-word highlighting options. There is no mention of training on this software.

RESULTS: The results from this adapted alternating treatments design, where students were exposed to alternating conditions (i.e., text plus audio alone, and highlighted text plus audio) indicated that text-to-speech can provide students with moderate intellectual disabilities improved independent access to content housed in both narrative and informational text. However, the results also suggest that highlighting did not contribute markedly to the comprehension or reading ability of students with moderate intellectual disabilities.

Erin K. Bone & Emily C. Bouck (2017) Accessible text-to-speech options for students who struggle with reading, Preventing School Failure: Alternative Education for Children and Youth, 61:1, 48-55

As students progress through school they spend more time reading to obtain information. Reading to learn can be a struggle for any student, but it tends to be a bigger obstacle for students with disabilities. Using text-to-speech applications and extensions is one way to assist students with disabilities who struggle to independently complete reading assignments. This article presents low-cost text-to-speech options that can be used to support struggling readers, including those with disabilities. Although many quality reading tools exist, this article provides options for teachers and other school personnel trying to balance the needs of students with the realities of school budgets.

Evmenova, A. S., & Regan, K. (2019). Supporting the Writing Process with Technology for Students with Disabilities. Intervention in School and Clinic, 55(2), 78–85.

Many students with learning disabilities (LD) and emotional/behavioral disorders (EBD) struggle with the writing process. Technology has shown to be effective in supporting prewriting, drafting, revising, proofreading, and publishing of written products. This article explains the use of one technology-based graphic organizer with embedded self-regulated learning strategies as well as universal design for learning (UDL) features that can be used to enhance the writing process for students with LD and EBD. Such technology categories as technology-based graphic organizers, word prediction, speech recognition, talking word processors, as well as multimedia and digital storytelling programs are discussed. All these technologies can improve the quantity and/or quality of student writing.

Izzo, M. V., Yurick, A., & McArrell, B., &  (2009). Supported eText: Effects of Text-to-Speech on Access and Achievement for High School Students with Disabilities. Journal of Special Education Technology 24(3), 9-20.

This quasi-experimental research study, published in a peer-reviewed academic journal, examined the effects of a text-to-speech screen reader program on the academic achievement of nine students in a high school resource room. The participants included seven high school students: one with a traumatic brain injury, one diagnosed with high-functioning autism, three with cognitive disabilities, one with an emotional disability and one with a learning disability. The open-source assistive technology program called CLiCk, Speak was used in this study to provide text-to-speech. The researchers providing training in how to use this text-to-speech software. A reverse research design was used and the text-to-speech intervention lasted approximately 14 weeks. A treatment fidelity checklist was used to measure adherence to the intervention project protocols.

RESULTS: The results of this study indicated the CLiCk, Speak translational support increased quiz performance with large effect sizes. Therefore, text-to-speech positively affects achievement supporting the use of text-to-speech as a translational resource for readers needing support to decode and comprehend content-area material.

Lange, A. A., McPhillips, M., Mulhern, G., & Wylie, J. (2006). Assistive software tools for secondary-level students with literacy difficulties. Journal of Special Education Technology, 21(3), 13.

This experimental research study, published in a peer-reviewed academic journal, examined the compensatory effectiveness of four assistive software tools (speech synthesis, spell checker, homophone tool, and dictionary) on literacy. For the purpose of this annotated bibliography only the speech synthesis results will be summarized. The participants included 93 students: 46 girls and 47 boys aged 14–15 years, who were at least one year behind in reading age as measured by the Group Reading Test II. The study adopted a two-factor mixed design and followed a pretest, training, posttest format. Participants were assigned to one of three groups: Assistive Software group, Microsoft Word Control group, and Full Control group. Groups were matched on IQ, reading ability, spelling ability, computer exposure, and socioeconomic status. The Assistive Software group used Read & Write Gold, Version 6 (2002), in conjunction with Microsoft Word; the Microsoft Word Control group used advanced tools in Microsoft Word; and the Full Control group accessed the tests through Microsoft Word but could not use the tools within the program. The Assistive Software group was trained on assistive software for use on the posttests, the Microsoft Word Control group was trained on assistive tools in the Microsoft Word software package for use on the posttests, and the Full Control group was not trained on software and had no assistance on the posttests. The effect of the speech synthesis assistive tool (Read & Write Gold) was assessed using a two-way mixed ANOVA with Pre-/Post-test performance as the within-subjects factor and Group as the between-groups factor.

RESULTS: Post-hoc tests comparing the Assistive Software group, Microsoft Word Control group, and Full Control group, respectively, revealed a significant improvement of approximately 8% for the Assistive Software group and no significant change for either the Microsoft Word Control group or the Full Control group. The Assistive Software group showed a significant improvement in reading comprehension using the speech synthesis tool, while the other two groups showed no improvement.

Meyer, N. K., & Bouck, E. C. (2014). The Impact of Text-to-Speech on Expository Reading for Adolescents with LD. Journal of Special Education Technology, 29(1), 21–33.

Texttospeech (TTS) technology holds promise as a compensatory tool for adolescents with learning disabilities in accessing grade-level expository text. A multiple-baseline-across-participants design examined the effectiveness of TTS on oral reading fluency, comprehension, and task completion time for two males and one female with reading disabilities in a Midwest junior high school. TTS did not affect students’ fluency, comprehension, or task completion time, although social validity interviews revealed that each student valued the independence and efficiency TTS provided. Students believed they comprehended fully, read more fluently, and finished the reading task more quickly with TTS than without it. Limitations and implications for future research are addressed in this article.

Pisha, B., & Coyne, P. (2001). Jumping off the page: Content area curriculum for the Internet age. Reading Online, 5(4).

This qualitative study was a formative evaluation of a digital prototype of a high school history book. The participants included 70 grade eleven students: 53 typical learners, 16 students with learning disabilities, and one student with a visual impairment.

RESULTS: This study revealed text-to-speech support was extremely popular with the readers with disabilities in the sample. Also, several students with no identified special needs reported they used the highlighting function of text-to-speech without the synthetic voice to support them. Another important finding was a number of students classified as typical learners chose to shut off the text-to-speech feature of the textbook.