Home > Reading and Spelling > Getting Up To Speed

Text-to-Speech Software for Reading

Reprinted with permission from the International Dyslexia Association quarterly newsletter, Perspectives, Summer 2007, vol. 33, no. 3, pages 11-16. (It's worth joining IDA just to get Perspectives. Their website is http://www.interdys.org.)

by Kenneth Elkind and Jerome Elkind
Students who are poor readers because of learning disabilities or attention disorders often read slowly, reread passages,
struggle to decode unfamiliar words, and interrupt their reading frequently to recover from fatigue and stress. Reading takes them much more time than it does their peers. As a result, they often do less well in school, and at some point in high school or college, many simply run out of time to complete their work.

Technology can help many of these students read more effectively and efficiently, allowing them to circumvent their reading difficulties. One key technology for providing reading assistance is text-to-speech (TTS) software. Assistive reading software uses TTS to speak the text as it is shown on the computer screen. In addition, the words are usually highlighted as they are being spoken, reinforcing the correspondence between the printed words and how they sound.

This article is about TTS software products designed to assist people with learning disabilities and attention disorders in their reading. It also touches briefly on the application of the technology to writing. The types of products available are discussed as are the reading tasks for which they are suited. We consider only products that run under the Windows operating system, that provide a synchronized presentation of speech and text, and that are mainly designed to read aloud entire documents. For information on the application of TTS to remedial instruction, see "Computer Assisted Instruction: Successful Only with Proper Implementation" and "Choosing Software Gems to Improve Children's Reading" in this issue and also Wise and Olson (1995).

BENEFITS OF TTS SOFTWARE

Speed

The most consistent benefit of TTS text-reading technology is that it enables slow readers to read faster. In studies of postsecondary students whose primary diagnosis was learning disabilities (Elkind, Black, & Murray 1996; Elkind, 1998) and attention disorder (Hecker, Elkind, Elkind, & Katz, 2002), almost all students whose unaided reading speed was less than 150 words per minute increased their reading speed by using the TTS technology. However, students whose unaided speed was above 200 words per minute were slowed by the TTS technology. A typical result from this set of studies is in Figure 1 (Elkind, 1998), which shows how the change in reading speed depended upon the students' unaided speed. The reason for this is straightfor­ward. The students set the TTS software to speak at normal con­versational speed, between 150 and 200 words per minute, regardless of their reading speed. Very slow readers increased their reading speed substantially to between 150 to 200 words per minute. Fast readers were slowed to the speaking speed.

Endurance

Student participants in our studies have repeatedly reported that TTS technology makes reading less stressful, less tiring, and more sustainable for longer periods of time. We observed this effect with a few students in our 1993 middle school study (Elkind, Cohen, & Murray, 1993). Our studies with college stu

dents confirmed this finding using structured interviews and questionnaires (Elkind, et aL, 1996; Elkind, 1998).

93% of the students with learning disabilities reported that reading was easier, less stressful, and less tiring.

91 % of students with learning disabilities said that they were able to increase the time that they could sustain attention to reading before their attention wandered or they needed a break.

The average duration of sustained reading reported by stu­dents with attention disorders increased about 60%, from 30-40 minutes to 50-60.

The combined effect of faster reading speed and longer read­ing durations can result in a dramatic increase in the amount of material that a slow reader can read in an extended reading ses­sion of several hours. Some slow readers saw improvements in the number of pages read by factors of 2 or 3.

Comprehension

A number of studies have shown that TTS improves compre­hension of many but not all students who are poor readers. Montali and Lewandowski (1996) and Elkind, Cohen, and Murray (1993) obtained this result with middle school students, as did Disseldorp and Chambers (2002) with ninth­grade students. Higgins and Raskind (1997); Elkind, Black, and Murray (1996); and Elkind (1998) obtained similar results with college students. It is important to note the finding of Montali and Lewandowski (1996) that merely presenting material orally, without a simultaneous visual presentation, did not change the comprehension abilities of poor readers. They surmised that without the visual presentation, the reader's attention is more likely to wander.
Leong (1995) pointed out the importance of determining the conditions under which TTS works well and the reading deficits of the students, reading materials, reading processes, and other variables that are most amenable to this technology. This same point was made by Disseldorp and Chambers (2002). A partial answer to this question was provided by Higgins and Raskind (1997); Elkind, Black, and Murray (1996); and Elkind (1998) who observed that students with poor unaided comprehension obtained greater improvements in comprehension than did stu­dents with good unaided comprehension. In fact, many of those with good comprehension actually experienced lower compre­hension scores. Figure 2 from our 1998 study is a typical result.

Figure 2. Effect of unaided comprehension on changes in comprehension from use of TTS. Nelson Denny Test used for determining comprehension standard scores.

The comprehension results are more scattered than the read­ing speed results discussed earlier. Scatter is measured by the correleation coefficient, r, which is -0.57 for comprehension versus -0.89 for speed. Therefore, unaided comprehension, by itself, is not an adequate predictor of the comprehensive bene­fits of TTS. Elkind, Black, and Murray (1996) and Elkind (1998) identified two other factors as being important for predicting
comprehension benefit: 1) the ability to attend simultaneously to both the auditory and visual presentation and 2) good auditory comprehension. The ability to attend to both presentations with­out interference allows a student to take advantage of the essen­tial contribution of the technology: simultaneous auditory and visual presentations. Students with good auditory comprehen­sion can be expected to obtain more information from the audi­tory presentation, which in turn should aid comprehension. This expectation was confirmed by their finding that students who scored high on the WAIS Vocabulary test (a test of oral vocabu­lary), and who therefore had a good understanding of the spo­ken words, experienced greater comprehension improvement than did students who had low scores.

Sustaining Attention

In our study of students whose primary diagnosis was attention disorders (Hecker et ai., 2002), we found that TTS technology reduced the frequency of distractions experienced in extended reading sessions. The median reduction in distractions ranged from 50% to 65% depending upon the experimental condition. Eighty percent of these students reported in structured questionnaires that they could concentrate longer, their attention wandered less, they were distracted less, and they skipped less material when using TTS technology. Students reported that thE duration of sustained reading increased from 32 to 52 minutes under one experimental condition and from 40 to 64 minutes under a second condition, both more than a 60% increase. This ability to sustain attention more consistently and for longer times led to an increase of 44%-54% in the number of pages that could be read per minute in extended reading sessions.

Writing

Although not the focus of this article, it is important to point out that TTS can also help students with writing. Raskind and Higgins (1995) found that college students using TTS technology to monitor their typing and to review their writing could locate a greater number of errors in their writing than they could when the text was read out loud by another person or when the worked without assistance. Overall, students found 36% of the errors in their writing with TTS technology versus 25% whe working unaided.

... continued in Part Two