What is “speech recognition technology”?
It is a relatively new word processing software. Rather than typing in your words, you speak into a microphone and the text appears on screen. “This is word processing using another input method,” says Follansbee. Or as a speech recognition website says, it is “computing out loud.” For users with physical disabilities, speech recognition also permits control over the entire computer.
How is it used in schools?
Speech recognition technology can accommodate a diverse group of students-who for a variety of reasons have difficulty writing. The software allows students with physical disabilities that prevent them from keyboarding as well as those with learning disabilities to compose sentences and edit their work.
Do users speak normally, as if in conversation?
Almost. There are two kinds of speech recognition technology. The first, called “discrete” requires the user to speak slowly, allowing the software to recognize your voice word by word. The newer type, called “continuous speech” allows the user to speak rather fluently, but it is far from the kind of regular speech you would use informally. One has to enunciate very clearly and avoid conversational “placeholders” like “um.” After the words appear on screen, the user can edit or change the text.
Who uses it?
People with physical disabilities have been using speech recognition for years. A high-profile example is Christopher Reeves, who has used the technology since he was injured in a horse riding accident. With the advent of continuous speech technology, the speech recognition industry is marketing their products to the general population. Reduced prices and availability have made it more available to others with learning or writing difficulties. So speech recognition technology is used by a great variety of people. Some professionals use it like they might have used a dictaphone in an another time. Lawyers find it useful and are as a group very adept at it because they have been trained to make organized oral presentations. Researchers use it to record field notes or observations. Medical specialists often use the software (which can be trained to recognize complicated medical terminology) to efficiently capture clinical records.
Is it hard to learn?
This change—from keyboard to voice—can be compared to the transition we all made from pen and paper to the keyboard, that is, one that requires some instruction and time to master. Learning speech recognition technology occurs at three levels. Mastering the software is one part. A second need is learning how to talk so that the software understands you. But the larger piece of learning is converting your speaking style to writing style. Students will initially speak in an informal, conversational style. They have to learn to reformulate into more of a “writing style.” This new technology puts the emphasis on how you think.
How do you “train” the software?
The software must be trained to recognize your voice. The first step for a user is to go through a structured training session provided through the software. By reading some straightforward texts, you “train” the software to understand you. And as you use the program over time, the software “learns” when you correct its mistakes.
Are more schools using it?
More and more. Until a few years ago, very few students used speech recognition in schools and very few schools knew about it. Many doubted its effectiveness and feared the expense. Now, with speech recognition’s increased visibility, more parents and student advocates are asking about it. Schools have traditionally turned to other kinds of assistive technology, but speech recognition now is making inroads. Some proactive school districts are even taking the lead, developing in-house expertise on this technology so that they can begin to introduce the technology to students as the need arises and before the students’ difficulties cause them unnecessary distress. It’s an advantage that the software can be used by more than one person, and some students have been able to teach their peers how to use it.
Originally published on June 1, 2000